Aurora Galactica And Platinum Project Guide

Updated: June 7, 2026

This is the maintained top-level explanation of the broader Aurora / Platinum effort. It is meant to answer four questions at once:

• what the project is trying to prove
• how Platinum, game applications, ingestion, harnesses, and release lanes fit

together

• how the conformance program is improving Aurora and future games
• how Codex, OpenAI models, local CPU/browser compute, and persistent artifacts

are being used to build more conformant software over time

For quick orientation before opening or resuming a branch, start with CURRENT_PROJECT_STATE.md. For measured reference-analysis and runtime-keeper cycles, use LONG_CYCLE_KEEPER_PROCESS.md.

Top-Level Goal

The goal is to build a repeatable system for creating polished, inspired, reference-conformant arcade software with modern LLM-supported engineering.

For Aurora, that means making a game that is not merely playable, but that continues to move closer to the gameplay pressure, stage choreography, sound feedback, visual discipline, and long-run mastery shape of Galaga-style arcade play.

For Platinum, that means proving a reusable platform that can host more than one serious game without absorbing game-specific truth.

For the conformance project, that means turning external artifacts into measurable evidence, turning measurements into game and platform changes, and tracking the resource cost of each improvement so future work is increasingly automated, local-compute-driven, and evidence-backed.

The Five Separated Layers

The project should be understood as five related but separate layers.

The main discipline is that reusable capability moves down into Platinum, while game truth stays with the game and reference truth stays in ingestion artifacts.

Current Project State

Aurora / Platinum is in a post-1.4.0 production development posture.

• hosted /production is the stable public 1.4.0 family
• hosted /beta remains the authority-gated review lane for the next deliberate

candidate cycle

• hosted /dev is the forward-review lane for the next coherent improvement

bundle and currently tracks the 1.4.0.1 forward-review line; exact current build label is authoritative in hosted build-info.json

• local localhost is the active engineering lane
• main is the authoritative integration line and currently includes the

consolidated Aurora challenge movement grammar, Galaxy Guardians ingestion cleanup, refreshed conformance economics, public project guide, white paper, slides, dashboards, release-schedule spine, and review packet

• release authority is currently on this MacBook profile

Current release-family ownership is maintained in RELEASE_SCHEDULE_AND_ISSUE_SPINE_2026-06-07.md. Use that schedule to keep docs, GitHub issues, workstreams, and lane gates aligned.

The active direction is no longer "can Platinum host Aurora?" It can. The active direction is:

• raise Aurora's Galaga-like conformance in the areas that most affect player

experience, especially audio feedback and challenge-stage set-piece quality

• make conformance measurement more precise and more automated
• keep local CPU/browser harnesses doing as much assessment work as possible
• use Codex and model calls to design, implement, analyze, and explain better

local systems rather than replacing measurement with opinion

• make Galaxy Guardians and later games arrive through ingestion-backed,

game-owned conformance packages

• keep the current workstream ordering explicit through

PROJECT_WIDE_WORKSTREAM_ALIGNMENT_2026-06-07.md

Current Path.Plan

The current path after the June 7 hosted /dev publish is:

1. Aurora challenge-stage runtime quality first: the grammar and first-five

reference contracts are ready enough to drive implementation, but runtime set-piece quality remains weak.

1. Galaxy Guardians v1 slice completeness second: opening presentation,

score/result identity, replay/video readiness, Watch/Rival/persona reuse, and Platinum frame parity.

1. Safety and release gates third: auth lanes, score integrity, replay/storage

boundaries, Supabase RLS posture, logging/privacy limits, and public/private dashboard separation.

1. Generated documentation and dashboards throughout: every serious score,

cost, artifact, and recommendation should remain visible in the public project/Application Guide surfaces, not only hidden in repo artifacts.

1. Low-risk delight after the guardrails stay green: Arcade Music, Watch Mode,

Player Two UX clarity, Commentator callouts, and optional immersive full-window gameplay may continue when they remain safely scoped, reversible, measured, and do not pollute production scores.

1. High-risk delight last: YouTube high-score posting and externally hosted

replay publishing must wait behind explicit auth, consent, storage, moderation, token, and revoke/failure-mode review.

The immediate work priority is to turn the new analysis and evidence into real runtime quality. The current consolidation was valuable, but beta should wait for a visible gameplay/platform lift beyond documentation and evidence maintenance.

May 19 Sprite And Formation Learning

The latest local work made one important process correction: sprite conformance cannot be represented by a single static crop score.

The project now needs to keep these reads separate:

The current formation-readability pass produced a useful split result:

• settled rack spacing now passes in normal rendering and Reference Pixel Lab

mode

• later-stage entry bursts were reduced by staggering non-stage-1 arrivals
• stage-1 opening crossing still produces warning-level overlap, so the next

true conformance step is a reference-timed opening path scorer rather than more ad hoc constant tuning

This is the kind of learning that must carry into Galaxy Guardians and later games. New games should not enter the pipeline with only a nice-looking static sprite sheet. They should bring a target crop, runtime crop, gameplay-scale read, formation/rack readability check, temporal motion plan, and entry-path grammar before the game is called conformant.

Cross-Game Ingestion Automation Direction

The project is moving toward a repeatable ingestion loop where the model helps design and interpret local harnesses, but the durable evidence is produced by local CPU/browser/video/audio work whenever possible.

For Aurora, the current sprite/readability cycle proved that source definitions, runtime canvas output, gameplay-scale proportions, and motion choreography can diverge. For Galaxy Guardians and later games, the first-class ingestion package therefore needs this maturity ladder:

The next conformance-program improvement should make this ladder visible as a cross-game dashboard row so Aurora, Galaxy Guardians, and future game packs can be compared by ingestion maturity as well as by current gameplay score.

Current Conformance Read

The latest maintained score roll-up is in CONFORMANCE_METRICS_OVERVIEW.md and the release planning dashboard is in RELEASE_CONFORMANCE_DASHBOARD.md.

Current read:

• conformance economics roll-up: 8.7/10
• measured resource ledger: 904 runs, about 58,277s wall time, about

58,392s CPU time, and about 1.48GB current artifact accounting

• application artifact conformance: 7.46/10
• weakest application artifact row: impact-explosion-visual-feedback
• Aurora challenge-stage set-piece conformance: 4.3/10 current average,

with movement 4.4/10, graphics 4.5/10, novelty 3.9/10, target-video object fit 3.6/10, challenge-specialty authority 3.2/10, and zero release-ready challenge contracts

• Aurora challenge grammar readiness: 25/25 first-five group contracts are

reference-backed and average control readiness is 8.6/10

• challenge motion primitives: 10 primitives, 4 high priority, release

readiness still planning-only

• Galaxy Guardians long-surface/persona review: 7.0/10, with stage-arc

pressure 7.9/10, stage presentation 7.4/10, persona review utility 6.5/10, midrun survivability 6.0/10, and stage-band authority 6.7/10

• overall Aurora quality remains strong in broad release reads, but the

project should not let broad scores hide the challenge-stage, event-feedback, and second-game completeness gaps

Important interpretation:

• a 10/10 score means "maxed at current scorer resolution," not arcade-perfect

imitation

• confidence and resolution are part of the metric, because a broad or young

scorer can hide meaningful feel gaps

• a better scorer may temporarily lower a score while improving the project

because it exposes a truer gap

• the latest audio and challenge-stage work is a good example: stabilizing

browser audio capture, adding composite analysis windows, and then adding calibrated browser-reference gates exposed the remaining cue-fit problem, while the stricter challenge-stage scorer lowered the apparent score because it now asks a better question about stage-by-stage arrival, alien novelty, choreography, and bonus-opportunity readability

• recent challenge-stage passes improved the dedicated score from 4.5/10 to

5.1/10, then to 5.8/10 after adding measured wave/group identity. The latest no-fire reference-motion extraction pass intentionally tightened the read to 5.7/10 conformance and 5.6/10 interesting factor by measuring authored challenge motion without bullet-truncated paths. The current gap is no longer only "do challenge stages exist safely"; it is now stronger trajectory/reference matching, the stage-11 challenge-3 reference miss, active sprite-motion novelty, late Galaga reference labels, and bonus-readability probes.

Per-Game Status

How The Platform Helps The Conformance Effort

Platinum is not just a web shell around Aurora. It is becoming the reusable infrastructure that makes measured game conformance practical.

Platinum helps by providing:

• local, dev, beta, and production lanes with explicit release identity
• a stable shell and pack-selection path for comparing games without rewriting

the host

• a future sprite-mode surface that can compare reference-conformance lanes,

Aurora production themes, and future-game variants without mixing target measurement with shipped creative art

• an optional full-window/cabinet display mode that can enlarge the active board

for play and review while preserving game-owned mechanics, input mapping, and replay evidence

• shared browser harness utilities and controlled-clock execution
• a developer conformance dashboard that can be shipped read-only with release

lanes

• release documentation gates that force score, evidence, and resource updates

before serious candidates

• platform-owned services for score transport, replay surfaces, feedback,

docs, and future evidence viewing

• explicit external-service access contracts, starting with the Supabase Data

API grants/RLS contract for profile and leaderboard tables

• common boundaries so each game can own its own mechanics and conformance

truth

The platform's value grows when conformance work discovers a reusable need and that need becomes a clean Platinum extension point instead of being buried in Aurora-specific code.

Where Ingestion Fits

Ingestion is the front half of conformance.

The ingestion framework turns external evidence into structured game knowledge:

1. source discovery and provenance
2. clip/window selection
3. frame, motion, waveform, and spectrogram extraction
4. semantic event logging
5. confidence and uncertainty notes
6. metric/scorer definitions
7. runtime correspondence targets
8. game-owned implementation and harness plans

For Aurora, ingestion keeps improving the reference basis for Galaga-like quality: audio clips, stage openings, challenge timing, boss formation grammar, pressure windows, contact sheets, and visual look checks.

For Galaxy Guardians, ingestion is even more central because the game should not become "Aurora with different labels." Its first credible playable phases should come from Galaxian-style source windows, event logs, sprite/cue targets, and runtime correspondence checks.

The maintained Galaxy-specific bridge between ingestion and release review is:

• GALAXY_GUARDIANS_FIRST_CLASS_CONFORMANCE_PLAN.md
• npm run harness:check:galaxy-guardians-first-class-conformance

For future games, ingestion should become the repeatable starting point rather than a manual design phase.

The Working Loop

The conformance process should keep cycling through this loop:

1. choose the highest-value conformance gap using score, confidence, user impact,

and resource economics

1. gather or refine reference evidence before subjective tuning
2. build or improve a local harness/scorer that can rerun the question
3. run local CPU/browser assessment long enough to get stable evidence
4. use Codex/model calls to design algorithms, review evidence, write code, and

summarize tradeoffs where that increases leverage

1. convert model-assisted insight into persisted repo logic whenever possible
2. change game or platform behavior only where the evidence supports it
3. rerun scorecards and dashboards
4. document score movement, confidence, cost, and next investment

This loop is deliberately "Karpathy-loop-like" in spirit: inspect concrete examples, improve the dataset and evaluator, make a small candidate change, run it, study failures, and fold the learning back into the system.

Resource Economics

The standing resource documentation is CONFORMANCE_ECONOMICS.md.

The standing self-critical work-block review is CONFORMANCE_INVESTMENT_RETROSPECTIVE.md. It is generated from challenge-stage history, quality score history, dashboard cost context, candidate sweeps, audio artifacts, and application artifact scores. Its job is to say plainly where score movement is real, where it is mostly measurement progress, and where Aurora is still not moving toward human-level conformance quickly enough.

The project currently reads as heavily local-first:

• local CPU/browser harnesses are doing the overwhelming majority of measured

conformance work

• Codex/OpenAI/model usage is essential for planning, implementation, synthesis,

and analysis, but is under-instrumented unless we log it manually

• any serious long-cycle work should be wrapped with npm run harness:measure

so we can compare resource cost against score movement

• after each large work block, run

npm run harness:analyze:conformance-investment-retrospective so the public docs and dashboard show the latest self-critical read

The desired pattern is:

• spend model/GPU-equivalent effort on strategy, harness design, code

generation, algorithm selection, and expert review

• push repeated measurement into local CPU/browser processes
• preserve every durable learning as code, metrics, reports, dashboards, or

artifacts

• prefer investments that improve both the game and the measuring system

Current Highest-Value Investment Areas

The dashboard currently keeps the detailed investment queue. At the strategic level, the important areas are:

1. audio identity, event feedback, and cue alignment
2. dedicated challenge-stage arrival, alien novelty, trajectory variation, and

high-bonus readability

1. regular alien entry geometry separation and path-family diversity
2. level arc and encounter shape
3. boss entry and formation grammar
4. visual look and feel, including start/attract typography and gameplay

readability

1. arcade cabinet frame, popup, help, scoring, leaderboard, and result surfaces
2. pressure curve precision and exact replay of known source windows
3. Guardians ingestion promotion from preview metrics into stronger

frame/audio/playtest evidence

1. Guardians score/progression/result identity so Platinum changes are

validated against two real games instead of one game plus one preview shell

The next major release gate should be judged not only by a higher overall score, but by whether the improved categories are the ones players actually feel.

Release Recommendation

Current recommendation:

• treat the refreshed public 1.4.0 family as the stable production line
• treat hosted /beta as the authority-gated review lane for the next

deliberate 1.4.0.1 candidate, not as an automatic mirror of local work

• treat hosted /dev and local topic branches as the place to assemble the

next coherent improvement bundle

• keep the larger 1.4.x family focused on arcade depth: audio feedback,

dedicated challenge-stage set pieces, alien entry, stage progression, boss/formation grammar, visual reference grounding, and measured player experience evidence

Short term:

• keep beta-facing release notes and readiness docs aligned with the current

1.4.0-beta.1 baseline and any new topic-branch improvements

• keep all candidate work wrapped in harness:measure
• keep dashboard, scorecard, public project page, project guide, release notes,

and strategic review aligned with the same current score read before the next beta candidate is cut

Latest local conformance response:

• stageTransition now uses a longer reference-backed window so the inter-level

moment is less abrupt.

• final ship loss now plays/logs gameOver before session export and keeps a

recording tail so the final audio does not disappear from captured evidence.

• wait-mode copy is simplified for the Aurora start screen while preserving the

game-picker hint and Guardians showcase content.

• boss first-hit feedback now uses compact damage sparks/ring/flash instead of

clunky full-destruction bursts.

• challenge-stage timing still passes, but challenge-stage arrival and pattern

novelty are now separated as their own conformance problem.

• sampled challenge stages preserve the safe Galaga-like no-shot/no-ship-loss

rule, but the dedicated set-piece score is still only 5.1/10.

• Stage 3, Stage 7, Stage 11, and Stage 15 now better separate their best

reference matches, while Stage 19 has a new runtime extraction path but needs late-reference labels before it can become a high-confidence conformance target.

Medium term:

• shape 1.4.0 around arcade depth: alien entry variation, challenge-stage

novelty, stage progression, boss/formation grammar, visual reference grounding, and measured audio feedback

Longer term:

• use 1.5.0 for shared video evidence and flight-recorder publishing
• use 1.6.0 for message-to-pilot and cabinet-surface polish
• reserve 2.0 for a credible multi-game Platinum milestone backed by

ingestion-driven game-owned conformance packages

Documentation System

The documentation should remain layered so readers can choose the right depth.

Maintenance Rules

Refresh this document when any of these change:

• release family or lane posture
• primary conformance priorities
• metric categories or interpretation rules
• alien/audio/stage/persona catalog entries
• ingestion's role in the project
• platform/application boundary strategy
• resource-economics policy
• status of Aurora, Galaxy Guardians, or a future game
• major hosted /beta pushes or beta-readiness decisions, which should also

refresh STRATEGIC_BETA_REVIEW.md

Before a serious /dev, /beta, or /production candidate, refresh:

npm run harness:build:conformance-metrics-overview
npm run harness:analyze:conformance-economics
npm run harness:build:release-conformance-dashboard
npm run harness:build:dev-conformance-dashboard
npm run build

Then update the human-readable docs if the generated artifacts changed the project story.

This is the current cross-thread alignment note after integrating the Aurora challenge movement grammar work, the Galaxy Guardians ingestion/conformance cleanup, and the hosted /dev publish from main.

Use this document when choosing the next work block. The goal is to keep Aurora, Galaxy Guardians, Platinum, personas, ingestion, documentation, and release gates moving as one program instead of as unrelated branches.

Current Lane State

• authoritative engineering source: main
• current source head: current main
• hosted /dev: current 1.4.0.1 forward-review build; exact label is

authoritative in hosted build-info.json

• hosted /beta: 1.4.0-beta.1+build.1013.sha.3cb0d08b.beta
• hosted /production: 1.4.0+build.894.sha.1dc23d8a
• release authority: MacBook-Pro, release-authority contract current
• hosted dev now includes the integrated Aurora challenge grammar artifacts,

Guardians cleanup artifacts, refreshed conformance economics, public project guide, white paper, slides, dashboards, release-schedule spine, and code-review packet

No beta or production publish happened during this consolidation.

The current release-family and GitHub-issue alignment is tracked in RELEASE_SCHEDULE_AND_ISSUE_SPINE_2026-06-07.md. Use that schedule to decide whether a work block belongs to 1.4.1, 1.4.2, 1.5.0, 1.6.0, 1.7.0, or 2.0.0.

Current Measured Reality

Priority Workstreams

Near-Term Execution Order

1. Keep main and hosted /dev clean after the current publish.
2. Start a focused Aurora challenge-stage runtime branch from current main.
3. Implement the first reference-spline movement candidate with explicit

visual-presence, phase/order, spacing, and scoreable-route gates.

1. Add a before/after stage video and contact-sheet comparison for that

candidate so humans can judge whether it actually looks better.

1. In parallel only where low collision risk, define the shared persona/Watch

adapter contract needed by both Aurora and Guardians.

1. Apply that contract to Guardians as the next v1 enabling step: Watch,

whole-run persona review, score ownership, and run summary should be game-owned but platform-driven.

1. Tighten Guardians opening slice using the existing source-backed artifacts:

WAIT, score table, rack cadence, combat feedback, top re-entry, and result state.

1. Refresh dashboards, conformance economics, and project docs after each

meaningful runtime candidate, not only after successful promotions.

1. Keep beta discussion blocked until hosted /dev contains at least one real

user-facing gameplay or platform-quality lift beyond documentation and evidence consolidation.

1. Preserve third-game work as ingestion and planning until Aurora challenge

quality and Guardians v1 both have better runtime proof.

Strategic Guardrails

• Do not treat high overall conformance scores as permission to ignore the

worst rows. The current user-visible weakness is challenge-stage spectacle, event feedback clarity, and Guardians game completeness.

• Do not ask the model for direct gameplay tuning before improving the

algorithmic assessment. The next useful Codex/model work should design analyzers, constraints, and candidate loops that local CPU/browser runs can execute repeatedly.

• Do not let Guardians borrow Aurora identities. Reuse Platinum infrastructure

and harness strategies, but keep Guardians game-owned scoring, visuals, audio, stage arc, and release truth.

• Do not move beta for docs alone. Documentation and dashboards can justify a

hosted /dev review, but beta needs an honest player-facing improvement or a clearly documented safety/release fix.

• Do not let third-game enthusiasm weaken the first two games. The third-game

path should validate the ingestion framework, not interrupt the Aurora and Guardians quality push.

Required Cross-Thread Checks

Run the smallest relevant set per branch, but these are the standing checks for the current integrated work family:

npm run build
npm run review:code:check
npm run publish:check:dev
npm run harness:check:portable-paths
npm run harness:check:challenge-movement-grammar
npm run harness:check:challenge-motion-primitives
npm run harness:check:challenge-motion-spec
npm run harness:check:galaxy-guardians-0-1-candidate
npm run harness:check:galaxy-guardians-long-surface-conformance
npm run harness:check:guardians-adapter-skeleton
npm run harness:check:pack-registry-boundaries

Use npm run harness:measure for long-cycle candidate sweeps so future planning can compare score movement against local CPU/browser and GPU-equivalent model/API effort.

This is the current future-looking release schedule for Aurora Galactica, Galaxy Guardians, Platinum, ingestion, personas, and conformance work.

Use this document as the release-family spine before opening a branch, accepting a beta candidate, or deciding where an issue belongs. Version numbers can be re-ordered when the actual bundle changes, but every active workstream should still map to one of these release families or be explicitly deferred.

Current Lane State

Verified June 7, 2026:

Current branch state:

• main is the authoritative engineering integration branch.
• release authority is currently held by macbook-pro.
• hosted /dev is expected to track the latest reviewed main dev publish.
• no beta or production publish is implied by this schedule.

Multi-Machine Allocation Model

The project can use multiple machines, but the release schedule remains the coordination layer. A branch should say both which release family it advances and which machine is expected to do the work.

Parallel-work rules:

• do not assign two machines to edit the same runtime surface without a named

integration branch and handoff note

• use iMac M1 for work that can run unattended or produce durable evidence
• use MacBook M4 for work that needs fast feedback, visual review, or release

integration

• push worker branches before switching machines or archiving sessions
• keep beta and production gated by release-authority.json, not by informal

machine habit

Release Family Schedule

Active Workstream Mapping

Issue Tracking Rules

GitHub issues live in sgwoods/Codex-Test1, not the public release host. The release host should remain a deployment mirror.

Each active issue should be assigned, in either labels or issue text, to:

• a release family: 1.4.1, 1.4.2, 1.5.0, 1.6.0, 1.7.0, or 2.0.0
• an area: aurora, guardians, platinum, ingestion, release, docs,

auth-storage, or third-game

• a proof type: runtime change, harness/evidence, docs, security,

manual review, or known deferral

Current issue audit from June 7, 2026:

• #242 is the latest triage item and should map to 1.5.0 because it is a

gameplay-export ingestion cycle for hosted-dev review.

• #201 maps to 1.4.1 because explosion, impact, damage, and immunity

feedback are directly tied to the weakest current application-artifact row.

• #202 maps to 1.6.0 unless a smaller reusable message/event-log primitive

is needed earlier for Watch/Rival/persona analysis.

• #203 maps to 1.4.2 because externally visible conformance dashboards need

Root/auth visibility rules before broader exposure.

• #140 maps to 1.4.1 because challenge stages must be treated and named as

bonus stages, not ordinary numbered stage progression.

• repeated movement, sound, pilot-score, screen-switching, high-score, and UI

polish issues should be deduplicated during the 1.4.2 issue hygiene pass before the next production promotion.

• unlabelled Google/security alert issues should be triaged separately from

game quality so operational noise does not obscure release blockers.

Gate Policy By Lane

Next Schedule Maintenance Actions

1. Keep this document linked from the hosted project guide, white paper, README,

product roadmap, and go-forward execution plan.

1. Before starting a branch, choose the release family it advances.
2. After each hosted /dev publish, update the lane state and any measured

score movements.

1. Before requesting beta, run an issue hygiene pass that maps or defers the

live open issues relevant to the release family.

1. Keep broad planning docs subordinate to this schedule: they can explain

strategy, but the schedule decides current release ownership.

This is the default Aurora workflow when development happens across more than one machine.

Core Model

• main is the only integration branch
• either machine may develop and push topic branches
• only one machine is the release authority at a time
• hosted /beta and hosted /production may only be published from the

current release-authority machine

The source of truth for release authority is:

• release-authority.json

Machine Roles And Work Allocation

The project may use more than one active development machine. The default division is based on what each machine is good at, not on ownership of a particular game.

Current practical rule:

• use the MacBook M4 for the highest-feedback gameplay and platform work
• use the iMac M1 for always-on or separable work that can produce durable

artifacts, branch commits, or GitHub issue updates

• every parallel thread must declare its release family from

RELEASE_SCHEDULE_AND_ISSUE_SPINE_2026-06-07.md

• every parallel thread must have a short handoff note before it is merged,

archived, or abandoned

Good iMac M1 assignments:

• Guardians ingestion and v1 evidence packets while the MacBook advances Aurora

gameplay quality

• overnight or multi-hour persona/watch runs with summarized artifacts
• gameplay-export ingestion cycles for hosted /dev review
• reference artifact portability checks and path sanitation
• GitHub issue deduplication and release-family labeling proposals
• documentation consistency sweeps after MacBook runtime work lands

Good MacBook M4 assignments:

• Aurora challenge-stage movement and visual-conformance runtime changes
• browser-visible UI, fullscreen, cabinet, score, and shell reviews
• heavy local conformance sweeps that benefit from the faster machine
• release-authority checks and hosted /dev publish from clean main
• integration of pushed worker branches from iMac M1 or future machines

One-Command Startup

After cloning the repo and installing the required system tools, the preferred startup path is:

mkdir -p "$HOME/Development"
cd "$HOME/Development"
curl -fsSL https://raw.githubusercontent.com/sgwoods/Codex-Test1/main/tools/dev/setup-machine.sh | bash

That installer:

• clones the repo into ./Codex-Test1 if needed
• reuses the clone if it already exists
• then runs npm run machine:bootstrap

The repo bootstrap itself:

• verifies node, npm, python3, gh, and Chrome
• verifies the repo remote and branch
• pulls origin/main if the repo is on a clean main
• runs npm install
• runs npm run build
• runs npm run local:resume
• writes the local-only .machine-profile.json
• reports live lane state, release authority, and next commands

For a read-only health check, use:

npm run machine:doctor

For a quick current-state summary, use:

npm run machine:status

Important path rule:

• prefer a non-iCloud local parent folder such as $HOME/Development
• do not keep the active Aurora clone inside an iCloud-managed folder
• do not rely on iCloud to synchronize .git state between machines

Branch Naming

Branch from main and use short-lived topic branches:

• codex/<machine-id>-<topic>

Examples:

• codex/imacm1-audio-polish
• codex/macbookpro-movement-traces

Recommended session pattern:

1. npm run machine:bootstrap or npm run machine:doctor
2. git switch main
3. git switch -c codex/<machine-id>-<topic>
4. make the smallest coherent change
5. run the smallest relevant gate set
6. commit and push the topic branch
7. merge back into main intentionally

Safe Switching Between Machines

Default rule:

• do not leave meaningful unpushed work on both machines at once without

documenting it

• do not use one shared working tree through iCloud or another sync layer
• before any multi-hour Codex run or when a chat is getting long, write a

durable context checkpoint:

npm run codex:checkpoint -- --label <short-topic> --plan "<current goal>" --next "<next concrete step>"

This writes:

• CODEX_CONTEXT_CHECKPOINT.md
• reference-artifacts/analyses/codex-context-checkpoint/latest.json

Codex Desktop does not expose a repository command that forces internal chat compaction. The recovery-safe replacement is to write this checkpoint, commit it when it captures meaningful state, then start a fresh Codex session from the restart prompt in CODEX_CONTEXT_CHECKPOINT.md.

Normal workflow:

1. start with npm run machine:bootstrap or npm run machine:doctor
2. if on main, pull before editing
3. if on a topic branch, rebase from updated main before publishing a

candidate

1. push before switching machines

When the repo is on a dirty main, bootstrap will not pull automatically. That is intentional. Clean up or branch off first.

Release Authority

Inspect the current authority with:

npm run release:show-authority

Hand off release authority intentionally with:

npm run release:claim-authority -- --machine-id <id> --label "<label>"

That command updates and commits release-authority.json, which makes the handoff visible in git history.

What A Non-Authority Machine May Do

Allowed:

• local development
• harness work
• docs and analysis work
• topic-branch pushes
• /dev preparation work on main

Not allowed:

• npm run approve:beta
• npm run publish:beta
• npm run promote:production
• npm run publish:production

Those commands are guarded and fail if this machine does not match release-authority.json.

Release Flow From The Authority Machine

Release-capable standard:

1. npm run machine:doctor
2. git switch main
3. confirm local main matches origin/main
4. run candidate checks
5. publish or promote from main

Current formal flow:

npm run approve:beta
npm run promote:production
npm run publish:check:production
npm run publish:production

Public sync is part of the same release discipline. verify:public must confirm:

• the Aurora project page
• the legacy alias
• the manifest JSON
• the rendered root sgwoods/public homepage card

Exceptional Case: Recovery-Safe Collaboration

If another machine holds meaningful uncommitted work that may be ahead of git, the normal workflow above is no longer sufficient.

That is when to use:

• RECOVERY_SAFE_COLLABORATION.md

Treat that document as the exception path, not the day-to-day default.

For the active migration away from iCloud-backed Aurora clones, also use:

• NON_ICLOUD_CLONE_MIGRATION_PLAN.md

Updated: June 5, 2026

This is the maintained index for game-specific conformance evidence across Platinum applications. It gives each game a durable catalog of:

• alien/enemy roles, names, activity, presence, and as-displayed conformance
• audio cues, target references, current evidence, and conformance posture
• stage-by-stage gameplay summaries and key conformance aspects
• testing personas used by platform and game harnesses

This catalog is a required high-priority artifact for every game that enters the ingestion pipeline. It should be updated whenever new extracted reference sprites, contact sheets, audio clips, stage windows, runtime evidence, or scorer outputs are promoted.

Scoring Interpretation

Catalog scores are x/10 at the current evidence resolution.

A 10/10 in another scorecard does not mean a row below is perfect. It means the specific scorer currently has no known measured gap.

Required Ingestion Artifact Set

For each newly ingested game, create and maintain these artifacts before the game is treated as a serious conformance target:

The canonical storage pattern is:

• human overview: this file
• source/media inventory: REFERENCE_MEDIA_INVENTORY.md
• game-specific evidence: reference-artifacts/analyses/<game-or-lineage>/
• promoted metrics: reference-artifacts/analyses/<metric-name>/latest.json
• runtime harnesses: tools/harness/

Recent Aurora sprite work adds one standing rule for future games: do not reduce alien or ship conformance to one static number. Track target likeness, live runtime crop, relative gameplay scale, formation readability, entry choreography, and temporal sprite motion separately. The new Aurora formation readability artifact lives at reference-artifacts/analyses/formation-readability/latest.json and should be used as the template for future rack/formation games.

For the second game and subsequent games, the catalog should also record which parts of the ingestion package are automatic, semi-automatic, or still human-reviewed. The desired direction is:

Aurora Galactica

Aurora is the active Galaga-style conformance target. It is an inspired game, not a pixel-copy project, so the strongest rows are those grounded in direct Galaga-family audio/timing/window evidence. Alien visual rows are currently less mature than the movement and stage-variation rows.

Alien And Enemy Index

Audio Cue Index

Stage Index

Persona Index

The current canonical persona profile is tools/harness/reference-profiles/persona-progression-correspondence.json. Future docs may expose beginner/intermediate/expert/professional aliases in the dashboard while keeping the historical novice/advanced/expert/professional ids stable for harness compatibility.

Distribution evidence is now promoted as a first-class persona artifact: reference-artifacts/analyses/persona-performance-distribution/latest.json. The current Aurora application guide renders a 30-run-per-persona score/stage chart and table from that artifact so one lucky or unlucky seed does not overstate the health of the persona ladder.

Galaxy Guardians

Galaxy Guardians is the Galaxian-style second-game preview and ingestion proof. It has stronger extracted visual reference artifacts than Aurora's alien visual catalog, but a much less mature playable/runtime conformance posture.

Alien And Enemy Index

Audio Cue Index

Stage Index

Persona Index

Galaxy Guardians now has game-owned deterministic personas, but they are still review tools rather than late-session authorities:

Galaxy Guardians First-Class Promotion Targets

The maintained Galaxy-specific promotion plan is:

• GALAXY_GUARDIANS_FIRST_CLASS_CONFORMANCE_PLAN.md
• GALAXY_GUARDIANS_LONG_SURFACE_AND_PERSONA_PLAN.md

Near-term first-class promotion targets:

• keep reference conformance at or above 7.7/10 while raising the

playtest-weighted score from 6.9/10 to at least the 7.0/10 compelling preview floor

• move formation/rack timing and motion pressure from proxy-tuned

6.2/10 reads toward 7.2/10 by promoting sprite-recognized timing and browser-reviewed motion feel

• move visual identity from 6.8/10 reference and 6.7/10 playtest toward

7.2/10 through human-reviewed sprite targets and gameplay-scale review

• move audio character from 6.4/10 toward 7.0/10 through human-listened

cue cleanup and cleaner isolated reference windows

• complete game-owned score/progression/result identity so the game validates

Platinum changes as a second real application, not only as a preview shell

Standing aggregate parity gate:

• npm run harness:check:galaxy-guardians-first-class-conformance

Windigo Invaders

Windigo Invaders is the planned Space Invaders-derived sibling game. It is not yet a runtime conformance target; it is currently an ingestion and planning target.

The important rule for this line is to avoid building an Aurora-shaped or Guardians-shaped shell from memory. The first meaningful output should be a strong source package and a Platinum-focused instantiation plan that separates host/platform ownership from game-owned truth.

Initial Intake And Planning Rows

Standing Planning Anchors

• WINDIGO_INVADERS_PLATINUM_INSTANTIATION_PLAN.md
• reference-artifacts/analyses/space-invaders-reference/README.md
• CLASSIC_ARCADE_INGESTION_FRAMEWORK.md

Platform Persona Contract

Platinum owns the persona harness substrate, not the game-specific truth behind each persona.

The platform should provide:

• stable generic persona IDs and aliases
• controlled seeds and deterministic clock execution
• scenario registration and result comparison
• dashboard display of persona outcomes
• distribution tables and charts across repeated seeded runs
• cross-game resource/economics accounting for persona runs

Each game should provide:

• which scenarios matter for each persona
• what success, failure, score, stage depth, and confidence mean
• how the generic personas map to that game's controls, threats, and scoring
• which behaviors are allowed to differ from the reference game
• which persona outcomes block a release gate

Maintenance Rule

When conformance work changes any alien role, audio cue, stage behavior, or persona assumption, update this file in the same commit as the harness or game change whenever practical. If the work is generated by long-cycle analysis, add the generated artifact path and note whether the row is high, medium, or low confidence.

The release path now treats catalog visibility as a gate, not a courtesy. Major artifact families that influence conformance decisions must appear in generated user-visible documentation or dashboard data before a lane publish check passes. Visual, audio, motion, entry, formation, or scoring changes should leave behind the artifact that proves what changed, what did not change, and what remains warning-level debt. Use this catalog for game-owned rows; use the release dashboard or conformance economics docs for release/resource rows.

Updated: May 30, 2026

Short restart note:

• GALAXY_GUARDIANS_RESUME_STATE_AND_NEXT_STEPS.md

Purpose

This plan turns Galaxy Guardians into a first-class conformance program on Platinum instead of treating it as only a preview card plus scattered harnesses.

Near-parity with Aurora does not mean the same score today.

It means Galaxy Guardians should have the same kinds of durable project surfaces:

• game-owned ingestion artifacts and provenance
• game-owned catalog rows for visuals, audio, stages, and personas
• game-owned runtime, scoring, and progression contracts
• game-owned conformance metrics and playtest review
• game-owned review gates plus explicit artifact-review checkpoints
• platform-boundary and two-game validation harnesses
• game-owned release framing, version tracking, and future candidate path

It also means Guardians should sit in the Platinum frame with the same class of platform-owned support Aurora already benefits from:

• pilot sign-in and signed-in game framing
• game-owned high scores, pilot records, and leaderboard views
• replay and video-capture/export surfaces
• bug-report and feedback transport surfaces
• Arcade Music, SFX, and mute/volume controls through the shared shell
• message, trophy, and other shared overlay surfaces that stay game-aware

instead of Aurora-shaped

As a standing rule, artifact-grounded evidence should lead the process. Human play/listen/look review is still allowed when the evidence package cannot yet settle a question cleanly, but it should be fallback confirmation rather than the primary baseline.

Aurora has also already paid for a large amount of reusable conformance machinery that Guardians should inherit rather than re-invent.

That shared machinery includes:

• promoted sprite-target pipelines and contamination guards
• temporal sprite-target extraction and sequence scoring
• waveform, spectrogram, cue-window, and audio-lab comparison flows
• hosted documentation, dashboards, release-note, and review-packet surfaces
• release-gate, review-disposition, and lane-verification discipline

Guardians should only diverge from Aurora's graphics/audio conformance tooling when the game grammar truly needs a different contract. The default is to reuse the same durable infrastructure with Guardians-owned targets, clips, score rows, and interpretation.

Current Baseline

The current maintained source of truth is:

• reference conformance: 7.6/10
• playtest-weighted conformance: 6.9/10
• long-surface and persona review: 7.0/10
• reference maturity: 7.2/10
• implementation gate coverage: 9.6/10
• public release readiness: 4.2/10

This means the Galaxy program is no longer missing machinery.

It already has a serious ingestion-backed evidence stack, but it is still short of first-class product parity because the highest-value remaining gaps are the ones a player feels immediately:

• opening WAIT/start-stage presentation and score-advance surfaces
• rack march cadence and left-right swarm motion
• explosion and hit/destruction visual feedback
• opening swarm palette families and stage-color progression
• moving starfield / scrolling background motion
• player readiness / missile-ready ship state
• reserve-ship and remaining-life icon fidelity
• level/stage flag graphics and progression markers
• bottom-pass-through wrap/return re-entry from top, not simple reappearance
• formation/rack timing
• lower-field motion pressure
• stage-five-and-beyond survivability and clear consistency
• sprite likeness at real gameplay scale
• audio character under live playback
• score/progression/end-state clarity
• game-owned review flow and release framing
• platform-frame parity across sign-in, scores, replay/video capture, bug

reports, music, and shared overlays

Three recent player-facing fixes are now in the maintained source:

• Guardians replay/video capture now persists into the local replay store with

game-owned metadata instead of being rejected as a preview-only start.

• Dive threats no longer drift visibly off the side of the board before the

intended bottom-exit/top-reentry loop.

• The live board now emits a recurring rack pulse, so representative gameplay

audio is less empty between shots, hits, and loss cues.

Three more high-value visible fixes are now in the maintained source:

• The live board now plays the formation-entry cue at the real

formation_entry_start event instead of only carrying a source-side cue id.

• Formation-mode aliens now have a visible pulse treatment rather than only

march motion plus static sprite rows.

• Special-value hit cases now surface score popups on the board, which makes

command-value and diving hits read more like intentional score events.

Public Slice vs. Deeper-Run Review

The hosted dev and beta lanes still expose a one-level visible public slice for Guardians.

This plan also tracks a deeper internal conformance layer built from repeated rack behavior, bounded stage-band escalation, and deterministic persona runs.

So when this document talks about stage five, later bands, or bounded late-loop pressure, it is describing the deeper runtime that the harnesses can already measure, not claiming that the current public preview already exposes multiple visible levels.

Do not treat later-band metrics as public-release marketing until the playable lane actually surfaces that depth.

The immediate visual baseline for the current public slice should therefore be treated as a first-order plan item, not a cosmetic afterthought.

Immediate Post-Merge Pickup

Once the incoming ship-sizing fix is merged, the next Guardians pass should resume in this order:

1. re-baseline the current public slice with the merged ship geometry and rerun

the opening-slice/render/motion harness spine before making more visual claims

1. preserve Aurora-built graphics and audio conformance machinery as shared

infrastructure, then swap in Guardians-owned evidence, targets, and runtime windows where needed instead of creating bespoke second-game pipelines

1. keep the now-improved visible one-level slice under guard, then move the

next polish pass to explosions, hit/destruction read, palette progression, and platform-frame fit

1. only after the opening slice looks and sounds authoritative, expand the

public playable depth beyond the current one-level mission_complete slice

1. use every resulting improvement as both a Guardians lift and a two-game

Platinum validation pressure test

First-Class Definition

Galaxy Guardians counts as a first-class conformance target when all of the following are true:

1. Its gameplay claims are grounded in committed Galaxian-family source

artifacts, not Aurora carry-over or memory-only tuning.

1. Its conformance story is readable from one maintained plan plus one

aggregate process gate, not only from many unrelated checks.

1. Its runtime owns score, progression, completion, loss, replay metadata, and

pilot-history identity as game-owned data.

1. Its harness family covers ingestion, visuals, audio, scoring, motion,

runtime flow, manual-review checkpoints, and platform boundaries.

1. Its next candidate path can be discussed as a game candidate, not only as a

platform curiosity.

Target Ladder

Strategy By Area

Required Harness Spine

The first-class Galaxy process should always be reviewable through this spine:

• ingestion and provenance
  • npm run harness:check:galaxian-reference-profile
• aggregate process parity
  • npm run harness:check:galaxy-guardians-first-class-conformance
• current reference and playtest posture
  • npm run harness:check:galaxy-guardians-reference-conformance
  • npm run harness:check:galaxy-guardians-playtest-conformance-review
• runtime completeness
• npm run harness:check:galaxy-guardians-runtime-slice
• npm run harness:check:galaxy-guardians-playable-preview
• npm run harness:check:galaxy-guardians-playable-preview should continue to

prove live-board starfield travel rather than only pack start/loss flow

• npm run harness:check:galaxy-guardians-game-owned-identity
• npm run harness:check:galaxy-guardians-one-level-completion
• npm run harness:check:galaxy-guardians-score-progression
• npm run harness:check:galaxy-guardians-threat-scoring
• opening-slice baseline
  • npm run harness:check:galaxy-guardians-opening-slice-baseline
  • npm run harness:check:galaxy-guardians-opening-slice-source-baseline
  • npm run harness:check:galaxy-guardians-opening-slice-frame-reference
  • npm run harness:check:galaxy-guardians-wait-launch-audio
  • npm run harness:check:galaxy-guardians-opening-slice-render-surface
  • npm run harness:check:galaxy-guardians-attract-score-surface
• longer-surface and persona review
  • npm run harness:analyze:galaxy-guardians-long-surface-conformance
  • npm run harness:check:galaxy-guardians-long-surface-conformance
• visual identity
  • npm run harness:check:galaxy-guardians-visual-readability
  • npm run harness:check:galaxy-guardians-sprite-grid-targets
  • npm run harness:check:galaxy-guardians-sprite-component-targets
• motion and rack pressure
  • npm run harness:check:galaxy-guardians-formation-entry
  • npm run harness:check:galaxy-guardians-movement-pacing
  • npm run harness:check:galaxy-guardians-stage-rank-pressure
  • npm run harness:check:galaxy-guardians-runtime-reference-movement
• audio identity
  • npm run harness:check:galaxy-guardians-audio-cue-targets
  • npm run harness:check:galaxy-guardians-audio-character
  • npm run harness:check:galaxy-guardians-audio-conformance-lab
• platform and multi-game boundaries
  • npm run harness:check:galaxy-guardians-replay-capture
  • npm run harness:check:gameplay-adapter-boundaries
  • npm run harness:check:platinum-pack-boot
  • npm run harness:check:compact-cabinet-rails

Artifact Review Loop

Automation is not enough yet for the last Guardians gaps, but the default should still be committed artifact comparison before human judgement. Each serious Galaxy review pass should first include:

1. rendered or captured side-by-side traces of rack settle, first dive,

flagship/escort pressure, and wrap/return timing against promoted windows

1. sprite/contact-sheet and attract/score-surface comparisons at gameplay

scale

1. cue-window, waveform, and spectrogram comparisons for shot, enemy-shot,

dive, hit, player-loss, and game-over moments

1. result-state and mission-complete artifacts so loss and completion language

are reviewable without preview-only placeholder messaging

Targeted human play/listen checks are still allowed when the artifact package does not yet settle the question, but they should be fallback confirmation rather than the primary baseline.

Opening-Slice Baseline Program

Because the current live public slice is only one visible level, the most important immediate misses are the first ones a player sees. The opening baseline program now already carries:

1. explicit opening audio-readiness and launch-audio confirmation
2. stronger first-seconds audible presence
3. tighter WAIT / start-stage headline and score-advance treatment
4. faster, more march-like left-right swarm cadence
5. moving starfield / scrolling background motion
6. missile-ready player-ship graphic/state
7. reserve-ship / ships-remaining icons
8. level/stage flag markers
9. bottom-exit ships visibly continuing through and re-entering from the top

rather than simply popping back in

The next phase after the opening-slice pass should therefore focus on:

1. using the new maintained stage-band authority inside the long-surface

review artifact to tune stage 3-5 pressure and stage 6-9 survivability/fairness instead of inventing fresh targets from plan prose

1. extracting richer behavior windows for dive timing, lower-field occupancy,

enemy-shot density, clear timing, and top-reentry continuity from those gameplay examples

1. browser-side listening review plus selective cue cleanup so the stronger

launch audio, recurring rack pulse, and combat presence become trustworthy public sound

1. a second WAIT / mission / score-table refinement pass only if hosted

review still shows obvious cabinet drift after the stronger combat, palette, and shell-fit pass

1. public-slice readiness work beyond the current one-level mission_complete

contract only after the improved opening and deeper behavior work survive hosted review

Each of these should be grounded in committed baseline artifacts from gameplay video, frame crops, contact sheets, palette extraction, timing traces, and other primary sources before runtime tuning broadens. The existing nenriki-15-wave-session long-session source should remain one of the main baseline references for background motion, wrap/return behavior, and sustained swarm feel.

The current human-readable opening-baseline summary lives in GALAXY_GUARDIANS_OPENING_SLICE_BASELINE.md, and the current machine-readable opening baseline now lives in:

• reference-artifacts/analyses/galaxy-guardians-identity/opening-slice-source-baseline-0.1.json
• reference-artifacts/analyses/galaxy-guardians-identity/opening-slice-frame-reference-0.1.json

Immediate Work To Start Now

1. Conformance-process parity

• add and keep the aggregate harness:check:galaxy-guardians-first-class-conformance

gate green

• make the core docs cite one Galaxy-first-class plan instead of only scattered

references

2. Game-owned operational identity

• complete score isolation so Guardians results do not silently mix with Aurora
• carry gameKey cleanly through score, replay, pilot, and result surfaces

3. Completion and result clarity

• replace preview-only completion messaging with real one-level completion and

loss result flow

• make the mission objective and end-state evidence explicit in runtime and

harness artifacts

4. Highest-value conformance lifts

• build the opening-slice baseline artifact package for WAIT, score table,

rack march cadence, explosion states, palettes, moving starfield motion, reserve ships, missile-ready state, stage flags, and bottom-pass-through top re-entry behavior

• reuse Aurora-proven graphics/audio machinery first, then promote

Guardians-owned target rows, contact sheets, cue windows, and runtime traces on top of that shared spine

• promote artifact-backed sprite targets and attract/score surfaces
• promote artifact-backed cue targets and runtime/reference audio pairs
• compare motion pressure after each measured timing change using captured

runtime/reference traces

5. Two-game validation value

• use Guardians as a standing second-game smoke for platform changes
• require multi-game validation whenever shell, score, replay, pilot, or pack

boundaries move

6. Platform-frame parity

• make Guardians a first-class consumer of pilot sign-in, pilot card, high

scores, leaderboard, and trophy surfaces

• make replay and video-capture/export flows game-aware for Guardians, not just

Aurora-shaped

• make bug-report and feedback transport preserve Guardians identity cleanly
• make Arcade Music, SFX, mute, and volume surfaces behave consistently for

Guardians through the shared shell

• use these parity surfaces as part of the standing two-game validation story,

not as optional polish after gameplay work

Relationship To Other Plans

• the active long working-block execution sequence is tracked in

GALAXY_GUARDIANS_8_HOUR_VISUAL_AUDIO_MOTION_PASS_PLAN.md

• the maintained longer-surface and persona review model is in

GALAXY_GUARDIANS_LONG_SURFACE_AND_PERSONA_PLAN.md

• the maintained game-arc, stage-band, and homage-variant grammar is in

GALAXY_GUARDIANS_STAGE_ARC_AND_HOMAGE_PLAN.md

• the earlier six-hour pass remains recorded in

GALAXY_GUARDIANS_6_HOUR_CONFORMANCE_PASS_PLAN.md

• branch-local gameplay-completeness work remains in

GALAXY_GUARDIANS_PLAYABLE_0_1_BRANCH_PLAN.md

• top-level release sequencing remains in PLAN.md and

PRODUCT_ROADMAP.md

• shared rules for how game evidence should enter Platinum remain in

CLASSIC_ARCADE_INGESTION_FRAMEWORK.md

This document is the maintained Galaxy-specific bridge between ingestion, runtime quality, review practice, and release planning.

Updated: May 13, 2026

Purpose

Galaxy Guardians can no longer be reviewed only as a one-level scout-wave peek in source, even though the hosted dev and beta lanes still present a one-level visible public slice.

The game still is not a full Aurora-depth release, but it now has enough runtime, scoring, presentation, and harness surface that we should judge it by the shape it really has:

• repeated rack play rather than one static board
• stage-band pressure that escalates and then clamps
• game-owned visual progression across later stages
• deterministic personas that can stress the game in repeatable ways

The maintained artifact for this review layer is:

• reference-artifacts/analyses/galaxy-guardians-identity/long-surface-conformance-0.1.json

The standing gate is:

• npm run harness:check:galaxy-guardians-long-surface-conformance

Current Measured Read

• reference conformance: 7.6/10
• playtest-weighted conformance: 6.9/10
• long-surface and persona review: 7.0/10
• stage-band pressure grammar: 7.9/10
• stage presentation progression: 7.4/10
• persona review utility: 6.9/10
• midrun survivability and clear potential: 5.8/10

The longer-surface program is now real, but the weakest remaining area is very clear:

• stage-five-and-beyond collision fairness
• later-band clear consistency
• stronger persona authority for late-session tuning

Public Slice Boundary

Hosted dev and beta currently present one visible mission-complete level.

This document governs the deeper internal repeated-rack and persona review layer we are using to shape the next depth step. It is not a claim that the current public preview already exposes multiple visible level cards or a finished later-game arc.

The next product step is to decide when to surface more of that depth publicly, rather than letting internal conformance terminology imply that it already has.

Before we expand the public depth, the current one-level public slice should also get its obvious baseline Galaxian-family cues into better shape: WAIT presentation, score advance table, swarm march cadence, explosion states, opening palettes, moving starfield motion, missile-ready ship state, reserve ships, level flags, and bottom-pass-through re-entry from the top instead of pop-in return.

Level Model

The current Guardians level surface should be judged by stage bands, not by expecting Aurora-style handcrafted stage identities already to exist.

These stage bands are runtime and harness bands, not separately surfaced public level labels in the current dev/beta slice.

That means opening-slice fidelity should improve in parallel with deeper-band work, not be skipped because the long-surface plan talks about later stages.

Persona Model

These personas are now game-owned and deterministic:

Current honest read:

• personas are now useful tuning partners
• personas are not yet authoritative late-band critics
• stage-five stress is measurable, but not yet consistently clearable

What This Pass Changed

This longer-surface tranche established three important corrections:

1. Life loss no longer rebuilds the whole rack.

The game now preserves board progress across deaths, which is much closer to the intended arcade shape and makes longer-surface review meaningful.

1. Later stages now have real presentation progression.

Stage themes, frame accents, and atmosphere no longer stay visually flat for the whole run.

1. Guardians now owns deterministic review personas.

The review layer can produce repeatable novice/advanced/expert/professional runs instead of borrowing Aurora-only persona assumptions.

Targets For The Next Multi-Hour Pass

1. Midrun fairness

Raise midrun survivability and clear potential from 5.8/10 toward at least 6.6/10.

Focus:

• lower-field dive collision fairness
• shot cadence under stage-five-and-beyond stress
• life-reset pacing after board-preserving respawn

2. Persona authority

Raise persona review utility from 6.9/10 toward at least 7.3/10.

Focus:

• clearer late-rack finishing behavior
• better stage-five route stability
• stronger difference between expert and professional outcomes

3. Longer-surface playtest credibility

Keep the evidence-weighted 7.6/10 reference score while raising the playtest-weighted 6.9/10 score past the compelling 7.0/10 bar.

Focus:

• opening-slice visible fidelity so the one public level looks more credibly

Galaxian-family before more depth is surfaced

• moving starfield/background motion grounded against the existing

nenriki-15-wave-session long-session source

• measured wrap/return behavior where bottom-exit ships continue through and

re-enter from the top rather than simply reappearing

• captured later-band runtime/reference traces
• promoted cue-window, waveform, and spectrogram comparisons after the latest

cue pass

• visible later-stage presentation evidenced through committed stage-band

artifacts rather than decorative claims

Required Companion Docs

• GALAXY_GUARDIANS_FIRST_CLASS_CONFORMANCE_PLAN.md
• GALAXY_GUARDIANS_STAGE_ARC_AND_HOMAGE_PLAN.md
• GALAXY_GUARDIANS_8_HOUR_VISUAL_AUDIO_MOTION_PASS_PLAN.md
• GAME_CONFORMANCE_CATALOG.md
• APPLICATIONS_ON_PLATINUM.md
• reference-artifacts/analyses/galaxy-guardians-identity/README.md

This plan is the maintained bridge between the current one-level public playable preview and the deeper internal repeated-rack game-quality work needed for Guardians to become a genuinely valuable second first-class game on Platinum.

Updated: May 19, 2026

Purpose

This document describes Galaxy Guardians as a game arc rather than only as a bag of harnesses.

It exists to answer four practical questions:

1. what the game should feel like from attract mode through loss or completion
2. how that feel should change by stage band as pressure rises
3. what conformance work should be sequenced first to make the game deeper in a

source-grounded way

1. what later homage variants may change without destroying the original

Galaxian-family flavor

This is a companion to:

• GALAXY_GUARDIANS_FIRST_CLASS_CONFORMANCE_PLAN.md
• GALAXY_GUARDIANS_LONG_SURFACE_AND_PERSONA_PLAN.md
• GALAXY_GUARDIANS_OPENING_SLICE_BASELINE.md
• CLASSIC_ARCADE_INGESTION_FRAMEWORK.md

Current Truth Boundary

Two things are true at once:

1. hosted dev and beta still expose a one-level visible public slice that

currently ends with mission_complete

1. the source repo already contains a deeper repeated-rack runtime and

stage-band conformance layer used for internal review, tuning, and future productization

So this plan distinguishes between:

• the current public-playable visible arc
• the deeper internal stage-band arc we are tuning toward
• the eventual homage-variant grammar that should sit on top of the validated

base game rather than replace it

Canonical Source Anchors

The current stage-arc and homage work should stay grounded in the committed Galaxian reference bundle:

• matt-hawkins-arcade-intro
• arcades-lounge-level-5
• nenriki-15-wave-session

These sources are already declared in:

• reference-artifacts/analyses/galaxian-reference/README.md
• reference-artifacts/analyses/galaxian-reference/source-manifest.json

Use them differently:

• matt-hawkins-arcade-intro for attract, WAIT, score-advance table, reserve

ships, flags, and the beginning of the first playable rack

• arcades-lounge-level-5 for a denser still-readable midgame pressure window
• nenriki-15-wave-session for moving starfield motion, longer-session rack

cadence, escort pressure, and bottom-exit/top-reentry behavior

Reuse Principle

Guardians should not be treated as a fresh tooling problem every time we return to it.

Aurora's recent visual/audio conformance work already established reusable infrastructure for:

• promoted sprite-target evidence and crop hygiene
• temporal target artifacts and sequence comparison
• cue contracts, waveform/spectrogram comparisons, and runtime audio labs
• hosted documentation, readable dashboards, and review/release packet flow

For Guardians, the default should be:

1. keep that shared machinery
2. replace Aurora-specific targets with Guardians/Galaxian-family targets
3. only invent new pipelines when the second game needs a genuinely different

contract

That keeps the second game from lagging only because the same support work is paid for twice.

Game Arc

1. Attract And Invitation

This is not a story cutscene. It is an arcade invitation.

The cabinet should communicate:

• high score and score advance table matter
• alien families are readable before play begins
• the game is about formation pressure, dives, and single-shot discipline
• the audiovisual identity is crisp, sparse, and readable rather than lush

Current biggest needs:

• stronger WAIT / headline treatment
• tighter score-advance table layout and phrasing
• better reserve-ship, flag, and HUD fidelity
• moving starfield and cabinet motion that already feels alive before play

2. Rack Establish

The first seconds of live play should establish the whole contract:

• the formation enters and settles into a marching rack
• the player gets one precise shot at a time
• the board is readable before real pressure begins
• the player ship looks armed and vulnerable

This is where conformance trust is won or lost fastest.

If the rack drifts lazily, the starfield is static, the ship lacks a ready cue, or the start surface reads as generic, the game feels off before the first dive.

3. Opening Pressure

After settle, the game should teach the basic Galaxian-family risk loop:

• scouts dive alone first
• the player learns the cost of a missed shot
• the lower field becomes dangerous without immediate chaos
• death hurts, but does not erase the board and destroy learning

The opening should feel fair, tight, and legible.

4. Command Pressure

Once the player understands the rack, flagship-and-escort moments should become the emotional center of the game:

• the board is no longer just a shooting gallery
• escorts and flagship dives become high-value/high-risk events
• enemy shots arrive often enough to force route discipline
• wrap and return behavior makes the lower field feel continuous

This is the moment where the game stops being "simple" and starts feeling distinctly Galaxian-family.

5. Sustained Session Pressure

A deeper run should not become a different game.

It should feel like the same grammar becoming denser:

• march cadence grows faster and more insistent
• dive timing compresses
• shot density rises
• stage presentation shifts enough to imply a real run, not a decorative skin
• expert and professional personas can survive longer by reading patterns, not

by exploiting non-arcade mechanics

6. Bounded Late Loop

The original family does not need a modern story ending to feel complete.

For conformance and future release work, the late loop should be treated as:

• a real high-pressure arcade state
• bounded and reviewable for harness purposes
• suitable for persona stress and fairness tuning
• still recognizably the same game, not a difficulty gimmick

The goal here is not infinite-system invention. The goal is authoritative pressure at the edge of the established grammar.

7. End States

There are three different "ends" we should keep separate:

1. current public slice end

Stage 1 clears and the current hosted preview completes the mission.

1. full-game arcade end

The primary canonical ending is player loss after a sustained run, with high score, stage reached, and run identity preserved.

1. variant or review cap end

A later homage or bounded review build may intentionally cap a session, but only if that cap feels like an arcade rule and not a story imposed from outside the original flavor.

Stage-Band Model

The current repo already has a stage-band review grammar. This table makes it readable as a game arc instead of only as internal tuning language.

Progression Of Work

We should advance the game in layers that improve both the live playable slice and the deeper conformance model.

Track 1. Opening-slice baseline

Best return against time and compute.

Scope:

• WAIT / score-advance table / headline surfaces
• moving starfield
• rack march cadence
• explosion states
• missile-ready ship state
• reserve ships
• stage flags
• bottom-exit/top-reentry visibility

Primary artifacts:

• attract/readiness crops
• contact sheets from matt-hawkins-arcade-intro
• long-session motion windows from nenriki-15-wave-session
• reference-artifacts/analyses/galaxy-guardians-identity/opening-slice-frame-reference-0.1.json
• reference-artifacts/analyses/galaxy-guardians-identity/opening-slice-source-baseline-0.1.json
• opening-slice render surface and attract-score surface artifacts already in

reference-artifacts/analyses/galaxy-guardians-identity/

Primary metric additions:

• explicit opening-stage presentation and HUD fidelity
• explicit rack-march and starfield-motion fidelity
• explicit reuse of Aurora-proven graphics/audio measurement machinery with

Guardians-owned targets and interpretations

Track 2. Pressure grammar and motion

This is where the game becomes deeper rather than merely prettier.

Scope:

• scout dive spacing
• flagship and escort pressure
• enemy shot cadence
• lower-field fairness
• top reentry timing and visibility
• life-reset pacing without full-board reset

Primary artifacts:

• runtime/reference traces
• promoted event windows
• longer-session motion captures
• stage-band timing summaries

Primary metric additions:

• stage-band pressure fidelity
• lower-field fairness and clear potential
• late-band persona divergence quality

Track 3. Score, loss, completion, and platform-frame parity

This is the highest-value structural layer after motion.

Scope:

• game-owned high scores and pilot records
• signed-in framing that feels as first-class as Aurora
• replay/video capture surfaces
• bug report and feedback transport
• music / SFX / mute / volume parity
• game-over and run-summary credibility

Primary metric additions:

• platform-frame parity
• run-summary and result-state fidelity
• two-game support parity between Aurora and Guardians

Track 4. Deeper repeated-rack authority

This is where the internal stage-band model becomes product-ready.

Scope:

• stage 3-5 and 6-9 evidence promotion
• stronger theme progression and pressure progression evidence
• late-loop clamp tuning
• expert and professional persona authority

Primary artifacts:

• stage-band captured traces
• persona distribution outputs
• later-session score and survivability summaries

Primary metric additions:

• stage-band coverage completeness
• persona authority score
• deeper-run public-readiness score once the product actually exposes it

Track 5. Homage variant framework

Variants should come after the base game is credible, not before.

The base game must first establish:

• recognizable Galaxian-family opening slice
• credible repeated-rack pressure grammar
• game-owned score/result identity
• first-class platform support inside Platinum

Only then should we invest in homage branches or selectable variants.

Homage Variant Grammar

Invariants To Preserve

Any true homage variant should preserve these:

• single-shot player discipline
• formation-rack establish before serious pressure
• march-like left-right cadence rather than floaty drift
• scout dives followed by flagship-and-escort command pressure
• visible bottom-exit and top reentry behavior
• score-advance literacy, reserve ships, flags, and high-score cabinet framing
• sparse, bright, cabinet-like audiovisual phrasing rather than cinematic wash
• a run that is fundamentally about survival, score, and pressure escalation

If a variant drops these, it may still be a fine game, but it is no longer a Guardians/Galaxian-family homage in the sense we want.

Safe Variant Levers

These are good later levers for novelty without losing flavor:

• faction naming and art direction within the same role grammar
• palette families by stage band
• cabinet-frame and marquee treatment
• alternate but equivalent attract/mission language
• slightly different dive curves and escort emphases
• audio timbre and phrase shaping around the same cue windows
• bounded alternate pressure presets for expert review modes
• optional capped-session variants if they still read as arcade rule sets

Unsafe Drift

These are likely to destroy the original flavor if introduced casually:

• Aurora carry-over such as capture/rescue, dual fighters, or challenge-stage

structures becoming the defining grammar

• rapid-fire or multi-shot player rules
• cinematic story cutscenes replacing score-table and rack-centered identity
• soft, floaty motion replacing march cadence
• decorative backgrounds replacing the live starfield and cabinet clarity
• wrap behavior that teleports or hides instead of visibly continuing the threat

Metric Strategy

We do not need a wholesale rewrite of the conformance model.

We do need to sharpen it in three ways:

1. add an explicit opening-stage presentation and HUD fidelity metric family

for Guardians

1. add an explicit platform-frame parity metric family so Aurora and

Guardians can be compared as first-class Platinum citizens, not just as isolated games

1. add a stage-arc coverage view so Guardians is judged by attract, opening,

escalation, sustained pressure, and end-state quality rather than by a flat overall score alone

That means:

• keep the current overall reference and playtest scores
• keep the current long-surface/persona layer
• add stage-arc readability on top rather than replacing the existing scores

Immediate Next Work

1. merge the incoming ship-sizing/runtime-size fix and rerun the opening-slice

artifact spine so the visible public slice is re-grounded on the corrected geometry

1. finish the opening-slice baseline so the public visible level looks clearly

Galaxian-family before we talk too much about later depth

1. reuse Aurora's promoted graphics/audio machinery for Guardians targets

wherever that shared contract still fits, instead of rebuilding equivalent second-game-only infrastructure

1. promote stage-band motion and pressure artifacts for stages 3-5 and 6-9
2. add a platform-frame parity scorecard row that compares Aurora and Guardians

directly

1. only after those steps, start designing homage variants from the preserved

invariant grammar instead of improvising novel mechanics too early

Current Top 10 Improvement Queue

This is the current ordered queue for Galaxy Guardians, balancing visible player value, conformance lift, and cost/compute discipline.

1. promote the opening-slice artifact pack for WAIT, score table, reserve

ships, flags, and player-ready/readiness cues from the committed intro sources

1. tighten rack march cadence and opening left-right swarm motion against the

measured opening windows

1. tune moving starfield motion and opening palette progression against the

long-session reference instead of decorative instinct

1. make bottom-exit / top-reentry behavior visibly continuous and timing-faithful

across the current public slice

1. improve explosion, hit, and destruction-state readability so combat feedback

feels more arcade-authentic at live scale

1. promote stage 3-5 pressure traces and tune the scout-to-flagship/escort

handoff into a clearer command-pressure arc

1. promote stage 6-9 traces and raise sustained-pressure fairness plus clear

potential for expert/professional personas

1. make Guardians first-class in the Platinum frame for sign-in, scores, pilot

card, replay/video capture, bug reports, and Arcade Music/SFX surfaces

1. strengthen game-over, mission-complete, high-score, and run-summary

credibility so the game reads as a real run and not only a preview slice

1. only after the above, define homage variants from preserved invariant

grammar rather than using novelty as a substitute for base-game authority

Relationship To The Current Runtime

This plan matches the current runtime and pack truth:

• stage theme progression already exists for stages 1, 3, 6, and 10+
• stage-rank pressure already exists as a bounded escalating runtime model
• the public playable slice still clamps maxPlayableStage to 1
• the internal harness model already supports deeper repeated-rack review

So the next job is not inventing a new game model from scratch.

It is making the current model readable, source-grounded, and progressively productized.

Updated: May 30, 2026

Purpose

This document narrows the current Galaxy Guardians conformance discussion to the first visible slice a player sees on hosted dev and beta.

That slice still decides whether the game feels like a real Galaxian-family cabinet or like a generic preview. Before deeper-stage claims matter, the opening needs to read correctly.

This is the human-readable companion to:

• reference-artifacts/analyses/galaxy-guardians-identity/opening-slice-frame-reference-0.1.json
• reference-artifacts/analyses/galaxy-guardians-identity/opening-slice-source-baseline-0.1.json
• reference-artifacts/analyses/galaxy-guardians-identity/opening-slice-render-surface-0.1.json
• reference-artifacts/analyses/galaxy-guardians-identity/opening-rack-motion-0.1.json
• reference-artifacts/analyses/galaxy-guardians-identity/score-progression-0.1.json
• GALAXY_GUARDIANS_FIRST_CLASS_CONFORMANCE_PLAN.md
• GALAXY_GUARDIANS_STAGE_ARC_AND_HOMAGE_PLAN.md

Current Visible Truth

Hosted dev currently exposes a one-level playable slice.

The strongest immediate fidelity questions are not deep later-band questions. They are opening-slice questions:

• does the WAIT / attract surface read like the source family
• does the score table feel like a real cabinet scoring surface
• do reserve ships, stage flags, and ready-to-fire cues feel native
• does the starfield move
• does the rack march
• do bottom exits continue into visible top re-entry

Those are the first seconds where trust is either earned or lost.

Source Anchors

1. Matt Hawkins arcade intro

Use matt-hawkins-arcade-intro for the pre-play identity layer:

• red title and mission language
• canonical headline: WE ARE THE GALAXIANS
• canonical mission line: MISSION: DESTROY ALIENS
• SCORE ADVANCE TABLE
• CONVOY / CHARGER score framing
• reserve ships and early cabinet HUD read
• first visible rack and player-ready state

The promoted windows already in the repo are:

• attract mission text: 0-10s
• score advance table: 10-35s
• formation entry start: 40-45s
• formation settle / stable rack: 45-55s

2. Nenriki 15-wave session

Use nenriki-15-wave-session for the live feel of the opening and early rack:

• moving starfield
• rack cadence once the board is alive
• visible player ship readiness near the bottom
• bottom-exit / top-reentry continuity
• longer-session proof that these cues are not only decorative attract elements

The promoted windows already in the repo are:

• stable rack read: 60-75s
• flagship / escort pressure: 90-135s
• enemy wrap or return: 105-150s

3. Galaxians example 12-minute session

Use galaxians-ex12mins as a corroborating repeated-start source:

• repeated attract/logo and score-advance-table confirmation across restarts
• early rack color-family and row-read checks
• restart cadence and quick early-score-strip review

This source is especially useful when the runtime is close but still feels "runtime tuned" rather than clearly source-family in the first minute, because it gives several repeated looks at the exact early-session surfaces players see.

Promoted Frame Windows

The opening no longer depends only on broad contact-sheet summaries.

The repo now carries a Guardians-owned promoted frame-window authority artifact:

• reference-artifacts/analyses/galaxy-guardians-identity/opening-slice-frame-reference-0.1.json

That artifact binds three concrete extracted windows into the current opening review path:

• matt-hawkins-arcade-intro/attract-mission-text
• matt-hawkins-arcade-intro/score-advance-table
• nenriki-15-wave-session/wrap-return-pressure

Each window already has a committed frame index, contact sheet, motion difference sheet, waveform, and spectrogram. The promoted Guardians artifact now also records representative sample-read timestamps so runtime layout and wrap-return work can cite specific frame windows instead of only plan prose.

What The Opening Should Read Like

Attract / ready surface

The opening should feel like an arcade invitation, not a modern tutorial panel.

The source-family read is:

• strong red title / mission language
• a visible score-advance surface before the first real rack matters
• sparse but legible HUD elements
• reserve ships and flags that read as cabinet state, not decorative UI
• a dark board already alive with motion

First playable rack

Once play begins, the player should read:

• a marching rack rather than a lazy drift
• a live starfield behind it
• a vulnerable single fighter with a visible ready-to-fire state
• a board that remains readable even before harder dives begin

Early threat continuity

Even in the current public slice, lower-field motion should imply a continuous threat model:

• ships can exit the bottom
• that threat continues rather than disappearing
• top re-entry is visible and part of the same pressure loop

Current Repo Read

Already present

• moving starfield is now visibly advancing on the live Guardians playable

preview board, not only in source-side/shared-board code paths

• starfield motion is now denser, brighter, and more directional instead of

reading like a near-static decorative scatter

• the Guardians-specific starfield renderer now honors the same intensity/speed

settings and render-debug reporting the rest of the platform expects

• the WAIT / preview mission block now preserves the canonical

MISSION: DESTROY ALIENS phrase inside a tighter Guardians-owned cabinet layout instead of a looser generic copy block

• score-table framing now uses CONVOY / CHARGER inside a compact aligned

wait-surface and preview-modal layout rather than a plain generic table

• launch from wait mode now explicitly confirms audio readiness instead of

silently depending on browser unlock behavior

• opening gameStart, formation pulse, and idle/wait cues now have a stronger

first-seconds audible presence

• the live board now actually plays the formation-start cue on

formation_entry_start, while recurring rack pressure uses the separate cadence lane instead of reusing the same cue for both jobs

• reserve ships and stage flags are visible
• a ready-to-fire missile cue is visible
• top re-entry now begins higher above the board and sweeps back in more

continuously instead of reading like a shallow pop-in

• hit and destruction feedback now uses a more sprite-like pixel-burst flash

rather than only line-burst rings

• special hit cases now also surface score popups on the live board, so diving

and command-value hits read more like deliberate score events

• preview palette ownership is still theme-aware, but stage one now preserves

the base game-owned scout / escort / flagship body colors more faithfully before later stage theming begins to widen

• formation-mode aliens now have a visible pulse treatment on the live board

instead of only static sprite rows plus rack motion

• stage-one formation rows now read with a stronger source-family ladder:

flagship yellow, escort red, upper scout violet, and lower scout blue bands instead of one flatter mostly role-owned opening swarm palette

• promoted frame windows now explicitly anchor the opening mission text,

score-table read, and wrap-return pressure instead of leaving those surfaces only at the level of broad source manifests

• the hosted docs now expose a stage assessment, a top-10 queue, and the full

deeper Guardians planning stack

Still partial

• the title line remains a Guardians-owned adaptation even though the primary

mission line now preserves the canonical MISSION: DESTROY ALIENS phrase

• the promoted frame windows now give the WAIT and score-table surfaces real

opening authority, but the current runtime still presents a three-row application-owned adaptation rather than a fully frame-matched cabinet copy

• rack march cadence now has its own object-track-derived opening contract, and

starfield/top-reentry now also have a promoted measured motion contract, but later review may still justify one more cabinet-window refinement pass if the live WAIT board keeps reading visibly looser than the promoted windows

• explosions and hit states now read better, and they now have frame-backed

source authority rather than only runtime improvement, but the live board still needs tighter matching against those windows

• palette progression is now more source-aligned in the opening slice, but the

stage-owned color story is still partial rather than fully measured

• the live pulse is now much easier to read, but it still needs one browser

side-by-side pass against the repeated early-session Galaxian sources before we claim the pulse treatment is fully source-tight

Still next

• one hosted side-by-side review pass to confirm the stronger combat, palette,

and shell-fit slice still reads correctly outside local preview

• browser-side listening review now that recurring rack pulse presence and

replay capture are both back in the live slice

• a second runtime WAIT, title, mission, and score-table refinement pass only

if that hosted review still shows obvious cabinet drift

• heavier stage 3-5 and 6-9 fairness work using the maintained long-surface

stage-band authority once the opening slice remains believable under hosted review

Why This Matters Before Deeper Stages

The deeper stage-band and persona work is still important, but the user is right to question it if the visible first level does not yet look convincingly Galaxian-family.

This opening baseline exists to keep the work honest:

• opening-slice authority first
• deeper stage-band claims second
• homage variants only after the base game is readable and trustworthy

Immediate Work Order

1. Preserve the opening-source baseline as a committed artifact and hosted

readable doc, not only as raw source manifests.

1. Review the improved opening slice side by side on hosted /dev, not only

on localhost.

1. Do the deliberate browser/listening pass now that the live slice has

stronger representative audio presence and working replay capture.

1. Do a second WAIT / title / mission / score-table pass only if hosted

review still shows obvious opening-window drift.

1. Once the opening slice holds, spend heavier effort on stage 3-5 and 6-9

pressure tuning through the maintained stage-band authority artifact.

Practical Conformance Meaning

If the opening slice becomes source-faithful enough, then later work on motion, midrun fairness, platform-frame parity, and homage variants has a believable foundation.

If the opening slice stays weak, later-stage scoring can look precise while the actual public game still feels obviously off in its first ten seconds.

Generated: 2026-06-09T13:28:39.710Z Commit: 925b2d8f Branch: codex/macbook-pro-1.4.1-stage7-object-track-keeper

Executive Summary

This is now a strict challenge-stage readout. The prior alien-entry score looked too healthy because it rewarded coverage, type labels, and broad stage signatures. That was useful harness progress, but it overstated the player-facing experience. This report follows the current project decision that challenging stages start at 1/10 interesting, 1/10 movement, and 1/10 graphical conformance until they earn credit through reference-grounded movement, active visual evidence, alien/stage novelty, and durable bonus-stage contracts.

Current result: 4.3/10 interesting factor and 4.3/10 challenge-stage conformance. Movement is 4.4/10, graphical conformance is 4.5/10, alien/stage novelty is 3.9/10, and player shot opportunity is 5.4/10. The strongest rule finding is that current probes show no enemy shots, no attack starts, and no ship losses during sampled challenge windows. The weakest player-facing finding is that current challenge stages are functionally safe but not yet fully credible Galaga-like bonus exhibitions: strict movement is 4.4/10, strict graphics is 4.5/10, alien/stage novelty is 3.9/10, player shot opportunity is 5.4/10, target-video object-track fit is 3.5/10, and sprite-motion correspondence is 6.18/10 with target timing status frame-labeled-segmented-reference-windows. Diagnostic legacy coverage was 6.8/10, which is why the old read was too generous.

Method

• Runtime challenge states were sampled through the browser-backed Aurora harness using challengeFormationState().
• Reference targets came from media-backed Galaga path labels and contact sheets.
• Existing path-family comparison supplied best-match vector scores against labeled Galaga challenge entries, but those broad scores are now retained as diagnostic coverage instead of the conformance score.
• Strict movement scoring compares runtime y-range, path length, turn count, reversals, lower-field share, and trajectory best-match against the selected Galaga challenge reference vector. It is capped by current temporal-measurement limits because the harness still samples summaries rather than full tracked choreography.
• Strict graphical scoring now includes active sprite-motion plus object-tracked runtime pixel/silhouette crops for flap state, phase coverage, visual family diversity, path-pose diversity, lit-pixel stability, and bounding-box variation. It remains capped until those object tracks are compared frame-by-frame to Galaga target crops, rotations, dive poses, capture/rescue transitions, and direct target crop sequences.
• First-pass target contracts now score group count, group path-family order, expected type order, and expected family order for any challenge that has a persisted media-backed contract. This is deliberately reported as a separate contract-fit read until target-video object tracking exists.
• Player shot-opportunity scoring samples plausible firing lanes through each challenge window so movement work can be judged by whether it creates learnable high-bonus routes, not only by broad movement shape.
• Challenge path-slot extraction suppresses player fire for challenge windows, so trajectory comparison measures authored alien motion instead of bullet-truncated player-score fragments.
• Safety is measured separately from interest: no shots/no kills is necessary, but it does not make a challenge visually conformant and contributes only as a guardrail.
• Prior 24-second evidence windows can include post-challenge normal play, so enemy bullets/attackers in those older windows are not treated as challenge-rule failures here.

Target Artifact Coverage

The broader Galaga target-artifact coverage read is 5.1/10 overall and 4.5/10 for challenge-stage target readiness. The important implication is not that Aurora lacks all grounding; it is that the currently ingested challenge-stage corpus is still early-stage heavy. Aurora now has media-backed windows for all tracked Galaga challenge stages, including Challenges 4-8. The bottleneck has moved from source acquisition to precision: each window needs five-group frame labels for entry side, path family, scoreable band, exit side, alien family, and perfect-bonus opportunity before direct trajectory scoring should rise.

Full target-artifact report: GALAGA_TARGET_ARTIFACT_COVERAGE.md and reference-artifacts/analyses/galaga-target-artifact-coverage/latest.json.

Stage Summary

Stage 3 / Challenge 1

Current score: interesting factor 4.2/10; challenge conformance 4.2/10. Movement 4.2/10, graphics 4.4/10, alien novelty 3.9/10, progression 3.2/10, player shot opportunity 5.6/10.

Legacy broad coverage score: 6.4/10. This is retained as diagnostic evidence only; it no longer counts as the player-facing conformance score.

Original target: First Galaga-style challenging stage: readable bonus set piece, no fire, no ship loss, upper-band mirrored entries, bee/butterfly line waves, visible arrival/peel-off.

Aurora current: first-challenge-peel / first-challenge-peel; lanes bee, bee, bee, bee, but, but, but, but; first-wave types bee, bee, bee, bee, but, but, but, but; visual families classic; strict movement 4.2/10, graphics 4.4/10, alien novelty 3.9/10, shot opportunity 5.6/10, target-contract fit 6.8/10.

Graphics read: Strict graphics score 4.4/10. Current visible family/type labels are present, Direct runtime-vs-Galaga target-crop comparator reads 5.97/10 overall and 3.86/10 for challenge specialty sprites; weakest challenge crop is challenge-mosquito at 3.68/10. Runtime sprite-motion hook observed 1 visual family/families, 3 path pose family/families, 6/6 animation phase buckets, both flap state(s), and 36 object-tracked silhouette track(s). Aurora runtime motion corresponds to 3 frame-labeled segmented-reference target row(s), average 6.2/10; phase-order average is 4.8/10. Weakest row bee-zako-pulse-pair at 5.42/10. Graphics remain capped at 5.1/10 until object tracks are compared as full temporal Galaga target-crop sequences, rotations, dive poses, and capture/rescue transitions.

Movement read: Strict movement score 4.2/10 against challenge-1-arrival-group-1: y-range fit 1, path-length fit 1, turn fit 1, object-track fit 0.24. Current probes now include runtime object tracks compared against CPU-extracted Galaga target-video object tracks where available. Legacy broad vector best-match was 8.5/10 against challenge-1-arrival-group-1; xRange 0.9441, yRange 0.7168, pathLength 0.9233.

Alien variation read: Strict alien/progression novelty score 3.9/10. Current stages expose labels, type mixes, direct specialty-sprite target-crop evidence, and 1 visual family/families, but this does not yet prove Galaga-like stage-by-stage introduction, fresh featured aliens, or memorable bonus-stage teaching moments. Opening wave exposes 2 type(s) and classic visual family labels. Group identity diagnostic: 5/5 wave type signatures and 3/5 path signatures; average within-wave spawn span 0.54s.

Group identity read: 5/5 wave type signatures and 3/5 path signatures; average within-wave spawn span 0.54s.

Target contract read: Target contract fit is 6.8/10 for Challenging Stage 1 (Levels 3-4): group count 1, runtime path-family order 1, declared path-family order n/a, type order 1, family order 1, object-track 0.24 via direct target-video tracks. This is a group/object-track read, not frame-perfect sprite identity recognition.

Shot-opportunity read: Shot-opportunity probe found scoreable targets in 61 sampled windows; 82% had a lane with 2+ targets, lane diversity was 1, and center-lane bias was 0.1.

Safety rule: enemy shots 0, attack starts 0, ship losses 0.

Critical gaps:

• Sprite-motion correspondence is 6.18/10: runtime flapping/cadence is visible, but target frame timing is still frame-labeled-segmented-reference-windows and weakest motion row is bee-zako-pulse-pair.
• Alien/stage novelty is only 3.9/10: the current type mix does not yet prove Galaga-like new alien introductions, memorable challenge-specific roles, or distinct bonus-stage learning patterns.

Next actions:

• Protect the first-challenge bee/butterfly line contract, then tune path length, turn count, and rack-slot precision against challenge-1 arrival and late-wave labels.
• Add contact-sheet comparison for first-visible frame, entry side, exit side, lane occupancy, and group timing so the next tuning pass can improve trajectory precision without subjective guessing.

Stage 7 / Challenge 2

Current score: interesting factor 4.2/10; challenge conformance 4.2/10. Movement 4.4/10, graphics 4.1/10, alien novelty 3.9/10, progression 3.2/10, player shot opportunity 5.5/10.

Legacy broad coverage score: 6.5/10. This is retained as diagnostic evidence only; it no longer counts as the player-facing conformance score.

Original target: Second challenge should feel denser and more novel than challenge 1 while staying nonlethal and non-shooting.

Aurora current: scorpion-cross-sweep / cross-sweep; lanes but, boss, rogue, bee, bee, rogue, boss, but; first-wave types but, boss, rogue, bee, bee, rogue, boss, but; visual families classic; strict movement 4.4/10, graphics 4.1/10, alien novelty 3.9/10, shot opportunity 5.5/10, target-contract fit 7.2/10.

Graphics read: Strict graphics score 4.1/10. Current visible family/type labels are present, Direct runtime-vs-Galaga target-crop comparator reads 5.97/10 overall and 3.86/10 for challenge specialty sprites; weakest challenge crop is challenge-mosquito at 3.68/10. Runtime sprite-motion hook observed 3 visual family/families, 3 path pose family/families, 6/6 animation phase buckets, both flap state(s), and 38 object-tracked silhouette track(s). Aurora runtime motion corresponds to 3 frame-labeled segmented-reference target row(s), average 6.2/10; phase-order average is 4.8/10. Weakest row bee-zako-pulse-pair at 5.42/10. Graphics remain capped at 5.1/10 until object tracks are compared as full temporal Galaga target-crop sequences, rotations, dive poses, and capture/rescue transitions.

Movement read: Strict movement score 4.4/10 against challenge-2-arrival-group-1: y-range fit 1, path-length fit 1, turn fit 0.9, object-track fit 0.5. Current probes now include runtime object tracks compared against CPU-extracted Galaga target-video object tracks where available. Legacy broad vector best-match was 8.7/10 against challenge-2-arrival-group-1; xRange 0.672, yRange 0.7129, pathLength 0.9196.

Alien variation read: Strict alien/progression novelty score 3.9/10. Current stages expose labels, type mixes, direct specialty-sprite target-crop evidence, and 1 visual family/families, but this does not yet prove Galaga-like stage-by-stage introduction, fresh featured aliens, or memorable bonus-stage teaching moments. Opening wave exposes 4 type(s) and classic visual family labels. Group identity diagnostic: 5/5 wave type signatures and 3/5 path signatures; average within-wave spawn span 0.53s.

Group identity read: 5/5 wave type signatures and 3/5 path signatures; average within-wave spawn span 0.53s.

Target contract read: Target contract fit is 7.2/10 for Challenging Stage 2 (Levels 7-8): group count 1, runtime path-family order 1, declared path-family order 1, type order 1, family order 1, object-track 0.5 via direct target-video tracks. This is a group/object-track read, not frame-perfect sprite identity recognition.

Shot-opportunity read: Shot-opportunity probe found scoreable targets in 47 sampled windows; 74% had a lane with 2+ targets, lane diversity was 1, and center-lane bias was 0.36.

Safety rule: enemy shots 0, attack starts 0, ship losses 0.

Critical gaps:

• Sprite-motion correspondence is 6.18/10: runtime flapping/cadence is visible, but target frame timing is still frame-labeled-segmented-reference-windows and weakest motion row is bee-zako-pulse-pair.
• Alien/stage novelty is only 3.9/10: the current type mix does not yet prove Galaga-like new alien introductions, memorable challenge-specific roles, or distinct bonus-stage learning patterns.
• Cross-sweep identity now lands on the expected Challenge 2 reference family; next work is trajectory precision and active visual novelty, not basic identity.

Next actions:

• Tune challenge 2 toward the denser mixed-novelty-line reference instead of relying on a generic cross-sweep.
• Score separate group identities so stage 7 is not just a slightly wider stage 3.

Stage 11 / Challenge 3

Current score: interesting factor 4.4/10; challenge conformance 4.4/10. Movement 4.4/10, graphics 4.6/10, alien novelty 3.9/10, progression 3.2/10, player shot opportunity 5.9/10.

Legacy broad coverage score: 6.9/10. This is retained as diagnostic evidence only; it no longer counts as the player-facing conformance score.

Original target: Third challenge should make the new visual family and boss-led novelty obvious, with larger sweep vocabulary and no attacks.

Aurora current: stingray-crown-hook-hybrid / hook-arc; lanes boss, rogue, but, bee, boss, bee, but, rogue; first-wave types boss, rogue, but, bee, boss, bee, but, rogue; visual families dragonfly; strict movement 4.4/10, graphics 4.6/10, alien novelty 3.9/10, shot opportunity 5.9/10, target-contract fit 7.2/10.

Graphics read: Strict graphics score 4.6/10. Current visible family/type labels are present, Direct runtime-vs-Galaga target-crop comparator reads 5.97/10 overall and 3.86/10 for challenge specialty sprites; weakest challenge crop is challenge-mosquito at 3.68/10. Runtime sprite-motion hook observed 3 visual family/families, 3 path pose family/families, 6/6 animation phase buckets, both flap state(s), and 40 object-tracked silhouette track(s). Aurora runtime motion corresponds to 3 frame-labeled segmented-reference target row(s), average 6.2/10; phase-order average is 4.8/10. Weakest row bee-zako-pulse-pair at 5.42/10. Graphics remain capped at 5.1/10 until object tracks are compared as full temporal Galaga target-crop sequences, rotations, dive poses, and capture/rescue transitions.

Movement read: Strict movement score 4.4/10 against challenge-3-arrival-group-1: y-range fit 1, path-length fit 1, turn fit 1, object-track fit 0.5. Current probes now include runtime object tracks compared against CPU-extracted Galaga target-video object tracks where available. Legacy broad vector best-match was 8.5/10 against challenge-3-arrival-group-1; xRange 0.5336, yRange 0.8672, pathLength 1.2258.

Alien variation read: Strict alien/progression novelty score 3.9/10. Current stages expose labels, type mixes, direct specialty-sprite target-crop evidence, and 1 visual family/families, but this does not yet prove Galaga-like stage-by-stage introduction, fresh featured aliens, or memorable bonus-stage teaching moments. Opening wave exposes 4 type(s) and dragonfly visual family labels. Group identity diagnostic: 5/5 wave type signatures and 3/5 path signatures; average within-wave spawn span 0.36s.

Group identity read: 5/5 wave type signatures and 3/5 path signatures; average within-wave spawn span 0.36s.

Target contract read: Target contract fit is 7.2/10 for Challenging Stage 3 (Levels 11-12): group count 1, runtime path-family order 1, declared path-family order 1, type order 1, family order 1, object-track 0.5 via direct target-video tracks. This is a group/object-track read, not frame-perfect sprite identity recognition.

Shot-opportunity read: Shot-opportunity probe found scoreable targets in 53 sampled windows; 89% had a lane with 2+ targets, lane diversity was 1, and center-lane bias was 0.28.

Safety rule: enemy shots 0, attack starts 0, ship losses 0.

Critical gaps:

• Sprite-motion correspondence is 6.18/10: runtime flapping/cadence is visible, but target frame timing is still frame-labeled-segmented-reference-windows and weakest motion row is bee-zako-pulse-pair.
• Alien/stage novelty is only 3.9/10: the current type mix does not yet prove Galaga-like new alien introductions, memorable challenge-specific roles, or distinct bonus-stage learning patterns.
• Dragonfly/boss-led identity now lands on the expected Challenge 3 reference family, but sprite-motion novelty and tracked Galaga challenge-3 path phases are not yet scored.

Next actions:

• Promote challenge-3 boss-led reference phases and score dragonfly visual novelty as animation, not only family label.
• Add motion windows for wing flaps, pulsing, dive/rotation silhouette, and challenge-only nonlethal arrival.

Stage 15 / Challenge 4

Current score: interesting factor 4.3/10; challenge conformance 4.2/10. Movement 4.3/10, graphics 4.6/10, alien novelty 3.9/10, progression 3.1/10, player shot opportunity 5.2/10.

Legacy broad coverage score: 6.9/10. This is retained as diagnostic evidence only; it no longer counts as the player-facing conformance score.

Original target: Fourth challenge should shift into long specialty serpentine arcs with obvious new color/family identity while staying nonlethal.

Aurora current: pink-serpentine-late / pink-serpentine; lanes boss, rogue, but, bee, bee, but, rogue, boss; first-wave types boss, rogue, but, bee, bee, but, rogue, boss; visual families galboss; strict movement 4.3/10, graphics 4.6/10, alien novelty 3.9/10, shot opportunity 5.2/10, target-contract fit 4.4/10.

Graphics read: Strict graphics score 4.6/10. Current visible family/type labels are present, Direct runtime-vs-Galaga target-crop comparator reads 5.97/10 overall and 3.86/10 for challenge specialty sprites; weakest challenge crop is challenge-mosquito at 3.68/10. Runtime sprite-motion hook observed 2 visual family/families, 3 path pose family/families, 6/6 animation phase buckets, both flap state(s), and 23 object-tracked silhouette track(s). Aurora runtime motion corresponds to 3 frame-labeled segmented-reference target row(s), average 6.2/10; phase-order average is 4.8/10. Weakest row bee-zako-pulse-pair at 5.42/10. Graphics remain capped at 5.1/10 until object tracks are compared as full temporal Galaga target-crop sequences, rotations, dive poses, and capture/rescue transitions.

Movement read: Strict movement score 4.3/10 against challenge-4-pink-serpentine-group-1: y-range fit 1, path-length fit 1, turn fit 1, object-track fit 0.33. Current probes now include runtime object tracks compared against CPU-extracted Galaga target-video object tracks where available. Legacy broad vector best-match was 8.9/10 against challenge-4-pink-serpentine-group-1; xRange 1.0153, yRange 0.5676, pathLength 1.6346.

Alien variation read: Strict alien/progression novelty score 3.9/10. Current stages expose labels, type mixes, direct specialty-sprite target-crop evidence, and 1 visual family/families, but this does not yet prove Galaga-like stage-by-stage introduction, fresh featured aliens, or memorable bonus-stage teaching moments. Opening wave exposes 4 type(s) and galboss visual family labels. Group identity diagnostic: 5/5 wave type signatures and 3/5 path signatures; average within-wave spawn span 0.24s.

Group identity read: 5/5 wave type signatures and 3/5 path signatures; average within-wave spawn span 0.24s.

Target contract read: Target contract fit is 4.4/10 for Challenging Stage 4 target-video object tracks: group count 1, runtime path-family order 0, declared path-family order n/a, type order 0, family order 0, object-track 0.33 via direct target-video tracks. This is a group/object-track read, not frame-perfect sprite identity recognition.

Shot-opportunity read: Shot-opportunity probe found scoreable targets in 58 sampled windows; 59% had a lane with 2+ targets, lane diversity was 1, and center-lane bias was 0.18.

Safety rule: enemy shots 0, attack starts 0, ship losses 0.

Critical gaps:

• Sprite-motion correspondence is 6.18/10: runtime flapping/cadence is visible, but target frame timing is still frame-labeled-segmented-reference-windows and weakest motion row is bee-zako-pulse-pair.
• Alien/stage novelty is only 3.9/10: the current type mix does not yet prove Galaga-like new alien introductions, memorable challenge-specific roles, or distinct bonus-stage learning patterns.
• Target-contract fit is only 4.4/10: the runtime group order, path families, type/family mix, or group count still misses the first-pass media-backed challenge contract.
• Pink-serpentine identity now lands on its first-pass late reference, but it still needs five-group frame labels, target sprite-motion evidence, and stronger player-visible novelty before it can claim maturity.

Next actions:

• Promote the Challenge 4 pink-serpentine window into five group labels and tune the runtime path so all groups keep a readable serpentine score lane.
• Add high-bonus readability probes so late-stage complexity stays learnable instead of becoming visual noise.

Stage 19 / Challenge 5

Current score: interesting factor 4.4/10; challenge conformance 4.4/10. Movement 4.6/10, graphics 4.7/10, alien novelty 3.9/10, progression 3.1/10, player shot opportunity 5.2/10.

Legacy broad coverage score: 7.1/10. This is retained as diagnostic evidence only; it no longer counts as the player-facing conformance score.

Original target: Fifth challenge should distinguish itself with pink/green cascade motion, alternating group identity, and stronger lower-field pass readability.

Aurora current: pink-green-cascade / pink-green-cascade; lanes rogue, boss, but, bee, boss, but, bee, rogue; first-wave types rogue, boss, but, bee, boss, but, bee, rogue; visual families galboss; strict movement 4.6/10, graphics 4.7/10, alien novelty 3.9/10, shot opportunity 5.2/10, target-contract fit 3.9/10.

Graphics read: Strict graphics score 4.7/10. Current visible family/type labels are present, Direct runtime-vs-Galaga target-crop comparator reads 5.97/10 overall and 3.86/10 for challenge specialty sprites; weakest challenge crop is challenge-mosquito at 3.68/10. Runtime sprite-motion hook observed 3 visual family/families, 3 path pose family/families, 6/6 animation phase buckets, both flap state(s), and 23 object-tracked silhouette track(s). Aurora runtime motion corresponds to 3 frame-labeled segmented-reference target row(s), average 6.2/10; phase-order average is 4.8/10. Weakest row bee-zako-pulse-pair at 5.42/10. Graphics remain capped at 5.1/10 until object tracks are compared as full temporal Galaga target-crop sequences, rotations, dive poses, and capture/rescue transitions.

Movement read: Strict movement score 4.6/10 against challenge-5-pink-green-cascade-group-2: y-range fit 1, path-length fit 1, turn fit 1, object-track fit 0.23. Current probes now include runtime object tracks compared against CPU-extracted Galaga target-video object tracks where available. Legacy broad vector best-match was 9.6/10 against challenge-5-pink-green-cascade-group-2; xRange 1.0344, yRange 0.6434, pathLength 1.9861.

Alien variation read: Strict alien/progression novelty score 3.9/10. Current stages expose labels, type mixes, direct specialty-sprite target-crop evidence, and 1 visual family/families, but this does not yet prove Galaga-like stage-by-stage introduction, fresh featured aliens, or memorable bonus-stage teaching moments. Opening wave exposes 4 type(s) and galboss visual family labels. Group identity diagnostic: 5/5 wave type signatures and 3/5 path signatures; average within-wave spawn span 0.25s.

Group identity read: 5/5 wave type signatures and 3/5 path signatures; average within-wave spawn span 0.25s.

Target contract read: Target contract fit is 3.9/10 for Challenging Stage 5 target-video object tracks: group count 1, runtime path-family order 0, declared path-family order n/a, type order 0, family order 0, object-track 0.23 via direct target-video tracks. This is a group/object-track read, not frame-perfect sprite identity recognition.

Shot-opportunity read: Shot-opportunity probe found scoreable targets in 45 sampled windows; 67% had a lane with 2+ targets, lane diversity was 1, and center-lane bias was 0.18.

Safety rule: enemy shots 0, attack starts 0, ship losses 0.

Critical gaps:

• Sprite-motion correspondence is 6.18/10: runtime flapping/cadence is visible, but target frame timing is still frame-labeled-segmented-reference-windows and weakest motion row is bee-zako-pulse-pair.
• Alien/stage novelty is only 3.9/10: the current type mix does not yet prove Galaga-like new alien introductions, memorable challenge-specific roles, or distinct bonus-stage learning patterns.
• Target-contract fit is only 3.9/10: the runtime group order, path families, type/family mix, or group count still misses the first-pass media-backed challenge contract.
• Pink/green cascade identity now lands on its first-pass late reference, but the current window still needs five group labels and stronger lower-field route readability.

Next actions:

• Promote the Challenge 5 pink/green cascade window into five group labels and tune lower-field pass timing against those labels.
• Score group-to-group alternation so cascade identity is not just a different path-family name.

Stage 23 / Challenge 6

Current score: interesting factor 4.2/10; challenge conformance 4.2/10. Movement 4.3/10, graphics 4.5/10, alien novelty 3.9/10, progression 2.7/10, player shot opportunity 5.5/10.

Legacy broad coverage score: 6.8/10. This is retained as diagnostic evidence only; it no longer counts as the player-facing conformance score.

Original target: Sixth challenge should emphasize green ladder rhythm and split exits.

Aurora current: green-ladder-split / green-ladder-split; lanes bee, but, rogue, boss, boss, rogue, but, bee; first-wave types bee, but, rogue, boss, boss, rogue, but, bee; visual families dragonfly; strict movement 4.3/10, graphics 4.5/10, alien novelty 3.9/10, shot opportunity 5.5/10, target-contract fit 4.6/10.

Graphics read: Strict graphics score 4.5/10. Current visible family/type labels are present, Direct runtime-vs-Galaga target-crop comparator reads 5.97/10 overall and 3.86/10 for challenge specialty sprites; weakest challenge crop is challenge-mosquito at 3.68/10. Runtime sprite-motion hook observed 2 visual family/families, 1 path pose family/families, 6/6 animation phase buckets, both flap state(s), and 30 object-tracked silhouette track(s). Aurora runtime motion corresponds to 3 frame-labeled segmented-reference target row(s), average 6.2/10; phase-order average is 4.8/10. Weakest row bee-zako-pulse-pair at 5.42/10. Graphics remain capped at 5.1/10 until object tracks are compared as full temporal Galaga target-crop sequences, rotations, dive poses, and capture/rescue transitions.

Movement read: Strict movement score 4.3/10 against challenge-6-green-ladder-split-group-1: y-range fit 1, path-length fit 1, turn fit 1, object-track fit 0.36. Current probes now include runtime object tracks compared against CPU-extracted Galaga target-video object tracks where available. Legacy broad vector best-match was 8.9/10 against challenge-6-green-ladder-split-group-1; xRange 1.0501, yRange 0.7984, pathLength 2.0261.

Alien variation read: Strict alien/progression novelty score 3.9/10. Current stages expose labels, type mixes, direct specialty-sprite target-crop evidence, and 1 visual family/families, but this does not yet prove Galaga-like stage-by-stage introduction, fresh featured aliens, or memorable bonus-stage teaching moments. Opening wave exposes 4 type(s) and dragonfly visual family labels. Group identity diagnostic: 5/5 wave type signatures and 1/5 path signatures; average within-wave spawn span 0.24s.

Group identity read: 5/5 wave type signatures and 1/5 path signatures; average within-wave spawn span 0.24s.

Target contract read: Target contract fit is 4.6/10 for Challenging Stage 6 target-video object tracks: group count 1, runtime path-family order 0, declared path-family order n/a, type order 0, family order 0, object-track 0.36 via direct target-video tracks. This is a group/object-track read, not frame-perfect sprite identity recognition.

Shot-opportunity read: Shot-opportunity probe found scoreable targets in 42 sampled windows; 81% had a lane with 2+ targets, lane diversity was 1, and center-lane bias was 0.17.

Safety rule: enemy shots 0, attack starts 0, ship losses 0.

Critical gaps:

• Sprite-motion correspondence is 6.18/10: runtime flapping/cadence is visible, but target frame timing is still frame-labeled-segmented-reference-windows and weakest motion row is bee-zako-pulse-pair.
• Alien/stage novelty is only 3.9/10: the current type mix does not yet prove Galaga-like new alien introductions, memorable challenge-specific roles, or distinct bonus-stage learning patterns.
• Target-contract fit is only 4.6/10: the runtime group order, path families, type/family mix, or group count still misses the first-pass media-backed challenge contract.
• Green-ladder split now lands on a Challenge 6 ladder/split reference; remaining work is fuller path length, lower reversal noise, and object-tracked group timing.

Next actions:

• Rebuild Challenge 6 around green ladder and split-exit timing until its measured vector hits challenge-6-green-ladder-split-group-1.
• Add frame labels for staggered ladder rungs, split exit side, and upper-band scoreability.

Stage 27 / Challenge 7

Current score: interesting factor 4.2/10; challenge conformance 4.2/10. Movement 4.5/10, graphics 4.6/10, alien novelty 3.9/10, progression 2.7/10, player shot opportunity 5.1/10.

Legacy broad coverage score: 6.9/10. This is retained as diagnostic evidence only; it no longer counts as the player-facing conformance score.

Original target: Seventh challenge should introduce a yellow diagonal fan with a memorable scoring lane.

Aurora current: yellow-diagonal-fan / yellow-diagonal-fan; lanes boss, bee, but, rogue, rogue, but, bee, boss; first-wave types boss, bee, but, rogue, rogue, but, bee, boss; visual families crown; strict movement 4.5/10, graphics 4.6/10, alien novelty 3.9/10, shot opportunity 5.1/10, target-contract fit 4.2/10.

Graphics read: Strict graphics score 4.6/10. Current visible family/type labels are present, Direct runtime-vs-Galaga target-crop comparator reads 5.97/10 overall and 3.86/10 for challenge specialty sprites; weakest challenge crop is challenge-mosquito at 3.68/10. Runtime sprite-motion hook observed 1 visual family/families, 1 path pose family/families, 6/6 animation phase buckets, both flap state(s), and 20 object-tracked silhouette track(s). Aurora runtime motion corresponds to 3 frame-labeled segmented-reference target row(s), average 6.2/10; phase-order average is 4.8/10. Weakest row bee-zako-pulse-pair at 5.42/10. Graphics remain capped at 5.1/10 until object tracks are compared as full temporal Galaga target-crop sequences, rotations, dive poses, and capture/rescue transitions.

Movement read: Strict movement score 4.5/10 against challenge-7-yellow-diagonal-fan-group-5: y-range fit 0.93, path-length fit 1, turn fit 1, object-track fit 0.29. Current probes now include runtime object tracks compared against CPU-extracted Galaga target-video object tracks where available. Legacy broad vector best-match was 9.5/10 against challenge-7-yellow-diagonal-fan-group-5; xRange 1.0468, yRange 0.538, pathLength 1.8102.

Alien variation read: Strict alien/progression novelty score 3.9/10. Current stages expose labels, type mixes, direct specialty-sprite target-crop evidence, and 1 visual family/families, but this does not yet prove Galaga-like stage-by-stage introduction, fresh featured aliens, or memorable bonus-stage teaching moments. Opening wave exposes 4 type(s) and crown visual family labels. Group identity diagnostic: 5/5 wave type signatures and 1/5 path signatures; average within-wave spawn span 0.23s.

Group identity read: 5/5 wave type signatures and 1/5 path signatures; average within-wave spawn span 0.23s.

Target contract read: Target contract fit is 4.2/10 for Challenging Stage 7 target-video object tracks: group count 1, runtime path-family order 0, declared path-family order n/a, type order 0, family order 0, object-track 0.29 via direct target-video tracks. This is a group/object-track read, not frame-perfect sprite identity recognition.

Shot-opportunity read: Shot-opportunity probe found scoreable targets in 31 sampled windows; 74% had a lane with 2+ targets, lane diversity was 1, and center-lane bias was 0.19.

Safety rule: enemy shots 0, attack starts 0, ship losses 0.

Critical gaps:

• Sprite-motion correspondence is 6.18/10: runtime flapping/cadence is visible, but target frame timing is still frame-labeled-segmented-reference-windows and weakest motion row is bee-zako-pulse-pair.
• Alien/stage novelty is only 3.9/10: the current type mix does not yet prove Galaga-like new alien introductions, memorable challenge-specific roles, or distinct bonus-stage learning patterns.
• Target-contract fit is only 4.2/10: the runtime group order, path families, type/family mix, or group count still misses the first-pass media-backed challenge contract.
• Yellow diagonal fan now lands on a Challenge 7 diagonal-fan reference; remaining work is stronger lower-field travel and object-tracked diagonal lane timing.

Next actions:

• Rebuild Challenge 7 around the yellow diagonal fan so the best match lands on challenge-7-yellow-diagonal-fan-group-1, not the finale label.
• Add a player-hit opportunity probe that rewards firing along the diagonal band rather than center-lane waiting.

Stage 31 / Challenge 8

Current score: interesting factor 4.4/10; challenge conformance 4.4/10. Movement 4.5/10, graphics 4.7/10, alien novelty 3.9/10, progression 3.1/10, player shot opportunity 5.5/10.

Legacy broad coverage score: 7/10. This is retained as diagnostic evidence only; it no longer counts as the player-facing conformance score.

Original target: Eighth visible challenge should act as a compact blue/purple late-loop capstone.

Aurora current: blue-purple-finale / blue-purple-finale; lanes rogue, boss, bee, but, but, bee, boss, rogue; first-wave types rogue, boss, bee, but, but, bee, boss, rogue; visual families stingray; strict movement 4.5/10, graphics 4.7/10, alien novelty 3.9/10, shot opportunity 5.5/10, target-contract fit 4.2/10.

Graphics read: Strict graphics score 4.7/10. Current visible family/type labels are present, Direct runtime-vs-Galaga target-crop comparator reads 5.97/10 overall and 3.86/10 for challenge specialty sprites; weakest challenge crop is challenge-mosquito at 3.68/10. Runtime sprite-motion hook observed 3 visual family/families, 3 path pose family/families, 6/6 animation phase buckets, both flap state(s), and 24 object-tracked silhouette track(s). Aurora runtime motion corresponds to 3 frame-labeled segmented-reference target row(s), average 6.2/10; phase-order average is 4.8/10. Weakest row bee-zako-pulse-pair at 5.42/10. Graphics remain capped at 5.1/10 until object tracks are compared as full temporal Galaga target-crop sequences, rotations, dive poses, and capture/rescue transitions.

Movement read: Strict movement score 4.5/10 against challenge-8-blue-purple-finale-group-4: y-range fit 1, path-length fit 1, turn fit 1, object-track fit 0.29. Current probes now include runtime object tracks compared against CPU-extracted Galaga target-video object tracks where available. Legacy broad vector best-match was 9.3/10 against challenge-8-blue-purple-finale-group-4; xRange 0.9897, yRange 0.7369, pathLength 2.5453.

Alien variation read: Strict alien/progression novelty score 3.9/10. Current stages expose labels, type mixes, direct specialty-sprite target-crop evidence, and 1 visual family/families, but this does not yet prove Galaga-like stage-by-stage introduction, fresh featured aliens, or memorable bonus-stage teaching moments. Opening wave exposes 4 type(s) and stingray visual family labels. Group identity diagnostic: 5/5 wave type signatures and 3/5 path signatures; average within-wave spawn span 0.21s.

Group identity read: 5/5 wave type signatures and 3/5 path signatures; average within-wave spawn span 0.21s.

Target contract read: Target contract fit is 4.2/10 for Challenging Stage 8 target-video object tracks: group count 1, runtime path-family order 0, declared path-family order n/a, type order 0, family order 0, object-track 0.29 via direct target-video tracks. This is a group/object-track read, not frame-perfect sprite identity recognition.

Shot-opportunity read: Shot-opportunity probe found scoreable targets in 38 sampled windows; 74% had a lane with 2+ targets, lane diversity was 1, and center-lane bias was 0.16.

Safety rule: enemy shots 0, attack starts 0, ship losses 0.

Critical gaps:

• Sprite-motion correspondence is 6.18/10: runtime flapping/cadence is visible, but target frame timing is still frame-labeled-segmented-reference-windows and weakest motion row is bee-zako-pulse-pair.
• Alien/stage novelty is only 3.9/10: the current type mix does not yet prove Galaga-like new alien introductions, memorable challenge-specific roles, or distinct bonus-stage learning patterns.
• Target-contract fit is only 4.2/10: the runtime group order, path families, type/family mix, or group count still misses the first-pass media-backed challenge contract.
• Blue/purple finale is now represented as its own runtime contract, but it still needs fuller path length and active sprite-motion evidence before it feels like a late-loop capstone.

Next actions:

• Refine the Challenge 8 blue/purple finale with fuller path length and compact late-loop timing.
• Promote challenge enemy active-motion scoring so visual novelty is measured through animation, not only family labels.

Plan To Improve

1. Treat the current strict scores as the release-facing truth: the broad coverage score is diagnostic only.
2. Build the challenge-stage target grammar: per-challenge group order, first-visible frame, entry side, exit side, path length, turn count, featured alien family, scoring window, perfect-bonus expectation, and result feedback.
3. Implement Stage 3 / Challenging Stage 1 first: top-right bee line, late top-left butterfly line, visibly longer upper-band sweep, clear peel-off exits, no combat grammar, and reference-matched duration/turn count.
4. Implement Stage 7 / Challenging Stage 2 as a different authored set piece, not just wider Stage 3: denser mixed novelty, crossing pattern, different entry side and exit side, readable scoring route.
5. Implement Stage 11 / Challenging Stage 3 with boss-led novelty and active animation evidence: featured alien role, flapping/pulsing/rotation windows, and a distinct reward read.
6. Continue the late-stage rebuild now that Challenges 4-8 have media-backed windows: preserve the new Stage 15, 19, 23, 27, and 31 contracts, then promote five group labels for each.
7. Prioritize Challenge 6 green-ladder and Challenge 7 yellow-fan precision work, because they now hit the expected reference families but still need longer path length, cleaner lower-field travel, and lower reversal noise.
8. Promote challenge-stage contact sheets, trajectory SVGs, active sprite-motion probes, and per-stage motion timelines into the Application Guide so the human review can see the actual delta, not only score text.

Success Criteria

• Raise challenge-stage interesting factor from 4.3/10 to 5.0/10 as the first honest beta-facing gate by implementing one visibly reference-like challenge, then toward 6.0/10 after Stage 3, 7, and 11 each have distinct authored contracts.
• Keep the separate target-contract read above 7.0/10 for Challenge Stage 1 while promoting contracts for Stages 7 and 11; group-contract success is useful but does not replace frame-level motion/graphics scoring.
• Raise movement conformance from 4.4/10 by increasing y-range, path length, turn count, and exit-side match against the Galaga challenge references.
• Raise graphical conformance from 4.5/10 by extending the object-tracked silhouette hook into Galaga target-crop sequence comparisons; do not inflate it from type labels alone.
• Raise player shot opportunity from 5.4/10 by creating lane-readable scoring windows for each challenge rather than incidental central-lane hits.
• Preserve 0 enemy shots, 0 enemy attack starts, and 0 ship losses during challenge windows.
• Convert the first-pass late-reference labels into five-group frame/object labels before treating the late challenge sequence as conformant.
• Extend sprite-motion scoring for challenge enemies so visual novelty becomes object-tracked pixel evidence, not only a phase/family hook.

Saved: 2026-06-05 Branch: codex/aurora-challenge-movement-grammar Baseline commit: 9cdd2d5ec Primary refreshed artifact: reference-artifacts/analyses/challenge-stage-conformance/latest.json

Purpose

This is the saved go-forward plan for Aurora's Galaga-style challenging-stage work. It preserves the current evidence-backed read before the next gameplay changes, stage by stage, and states the most specific work that can move each stage toward a major improvement.

The important current finding is narrow but useful: recent runtime work improved clock handling, spacing, and Challenge 2 identity enough to nudge the strict score from about 4.2/10 to 4.3/10. That is real headway, but not a major player-facing transformation yet. The large remaining gap is still target-video choreography, active alien visual novelty, and human-learnable perfect-score route design.

Naming Note

Runtime and historical artifacts currently use internal stage markers 3, 7, 11, 15, 19, 23, 27, and 31 for the eight tracked challenge windows. Human-facing documentation should avoid implying that these are ordinary levels. Going forward, use the pattern Challenging Stage A-B where the runtime mapping confirms the challenge occurs between regular stages A and B, while preserving the internal marker in evidence tables.

Until that mapping is cleaned up everywhere, this report uses:

Challenge N / internal stage marker M

Current Topline Read

Interpretation:

• Safety is the strongest rule: sampled challenge windows preserve no enemy shots, no attacks, and no ship losses.
• Legacy coverage is not the player-facing score. It says we have broad labels and path-family coverage, not that the challenges feel conformant.
• The current hard blocker is visual and temporal specificity: the target-video object-track fit is 3.6/10, novelty is 3.9/10, and sprite motion is still capped by frame-labeled segmented-reference windows.
• Candidate sweeps have produced useful evidence but no runtime keeper across all eight stages.

Saved Go-Forward Plan Set

1. Treat strict challenge-stage conformance as the release-facing truth. Broad coverage remains diagnostic only.
2. Keep safety as a hard guardrail: no enemy fire, no attacks, no ship loss during challenge windows.
3. Finish the motion grammar vocabulary so each challenge is defined as five groups with entry side, exit side, path family, timing, score lane, alien family, and result opportunity.
4. Make the first five challenges the near-term quality target, because they are the user-visible foundation and currently have the strongest grammar artifacts.
5. Do not promote a runtime candidate only because one proxy improves. Promotion must improve or preserve target-video fit, human-visible readability, human-perfect routeability, spacing, and identity.
6. Use synchronized target/current video and contact sheets as review artifacts before changing runtime tuning.
7. Add active sprite-motion scoring for challenge aliens: flap cadence, pulse, rotation/dive silhouette, and specialty alien phases.
8. Add persona-perfect route checks so a strong persona can actually hit the intended groups without hidden or unfair timing.
9. Feed spacing, lead-in continuity, and no-magic-appearance checks into strict challenge scoring instead of leaving them as separate guardrails.
10. Update Application Guide and project docs after each meaningful pass so the human-visible story matches the current artifacts.

Evidence Sources

• CHALLENGE_STAGE_CONFORMANCE_ANALYSIS.md
• CHALLENGE_SETPIECE_CONTRACTS.md
• GALAGA_CHALLENGE_VIDEO_REFERENCE_ANALYSIS.md
• GALAGA_TARGET_ARTIFACT_COVERAGE.md
• reference-artifacts/analyses/challenge-stage-conformance/latest.json
• reference-artifacts/analyses/challenge-stage-candidate-sweep-index/latest.json
• reference-artifacts/analyses/challenge-movement-grammar/latest.json
• reference-artifacts/analyses/challenge-motion-primitives/latest.json
• reference-artifacts/analyses/galaga-challenge-object-tracks/latest.json
• tools/harness/check-challenge-motion-profile.js

Current local runtime work behind this report:

• Delayed challenge-wave clocks no longer age while enemies are hidden.
• Reference-backed challenge paths now have side-aware lane offsets.
• Challenge motion/profile guardrails now cover delayed wave starts and spacing/bunching risk.
• The refreshed analyzer now reads Challenge 2 against the expected challenge-2-arrival-group-1 family instead of incorrectly matching Challenge 1.

Stage-By-Stage Assessment

Challenge 1 / Internal Stage Marker 3

Current score: 4.2/10

Target state:

• The first Galaga-style challenge should teach that this is a safe score exhibition, not combat.
• It should show clean bee/butterfly group entries, readable upper-band passes, peel-off exits, and a learnable route toward perfect bonus.

Current Aurora state:

• Runtime identity is now a credible first-challenge peel: bee and butterfly lanes, classic family labels, and no-shoot/no-kill safety.
• Target contract fit is the strongest of the current set at 6.9/10.
• The visible target-video fit remains weak at 3.2/10, so the stage is not yet moving like the baseline.

Differences and conformance gap:

• It has the right broad idea, but not enough frame-level timing, exact group spacing, or score-lane choreography.
• The current visual score is capped because active sprite motion is not yet compared as a target-crop sequence.

Most specific work to improve:

1. Promote full five-group frame labels for Challenge 1, including first-visible frame, entry side, exit side, and scoreable band.
2. Feed the new delayed-clock and spacing guardrail measurements into the strict scorer so current stability improvements are scored directly.
3. Re-run candidate search with hard bunching limits and reject candidates that hurt human-visible readability.
4. Add a persona-perfect route probe for Challenge 1 so the stage is judged by whether a skilled player can actually learn and clear the set piece.
5. Generate a synchronized target/current video from stage start to result screen for manual review.

Challenge 2 / Internal Stage Marker 7

Current score: 4.2/10

Target state:

• The second challenge should not feel like a wider Challenge 1.
• It should introduce denser mixed-family movement and a crossing pattern that is still safe and learnable.

Current Aurora state:

• This stage improved materially in the refreshed analyzer: the best match now lands on the expected Challenge 2 reference family.
• Shot opportunity is relatively strong at 6.0/10.
• Graphics remain weak at 4.1/10, so it is still not a strong visual spectacle.

Differences and conformance gap:

• The broad identity is now right, but the stage still lacks a crisp cross-sweep visual read and active sprite novelty.
• The candidate sweep had no keeper; the best prior candidate had only +0.1/10 human-visible lift.

Most specific work to improve:

1. Lock this as the first post-fix review target because it now has the best combination of correct identity and target-video fit.
2. Add group-by-group target labels for the crossing pattern and score separate group identities.
3. Strengthen graphics by measuring active alien animation and specialty poses during the crossing route.
4. Tune path spacing and phase offsets against the reference, not against a generic cross-sweep label.
5. Add a human-visible pass/fail guard for "not magic appearing" and coherent group entry-to-exit continuity.

Challenge 3 / Internal Stage Marker 11

Current score: 4.4/10

Target state:

• This should become a boss-led or new-family challenge with visibly different motion from Challenges 1 and 2.
• It should make specialty alien identity and active animation obvious.

Current Aurora state:

• This is currently one of the better measured stages: strict 4.4/10, target contract 7.2/10, target-video fit 4.7/10.
• The dragonfly/boss-led identity lands on the expected reference family.

Differences and conformance gap:

• The stage still gets only 3.9/10 novelty because alien and animation evidence is too label-based.
• The target phase timing is not yet strong enough to claim real Galaga-like motion.

Most specific work to improve:

1. Promote Challenge 3 boss-led frame labels and active sprite-motion targets.
2. Add direct scoring for flap, pulse, and path-pose phase order during the stage.
3. Build a target/current video review pair that highlights the boss-led entry and exit rhythm.
4. Tune lower-field score lanes so a skilled persona has a clearer perfect route.
5. Convert the current high target-contract fit into a runtime keeper only after target-video and human-visible gates improve together.

Challenge 4 / Internal Stage Marker 15

Current score: 4.2/10

Target state:

• This should be the first clearly late-stage feeling challenge: specialty serpentine arcs, more memorable path shape, and stronger color/family identity.

Current Aurora state:

• The stage is correctly named around pink-serpentine identity, but the measured contract fit is only 4.4/10.
• Candidate search had many attempts but no safe runtime keeper; prior evidence showed proxy gains that did not translate into target-video or human-visible gains.

Differences and conformance gap:

• The labels say "serpentine"; the current object tracks do not yet prove the target's serpentine motion.
• Shot opportunity dropped to 5.2/10, which means late complexity may be reducing fair perfect-score potential.

Most specific work to improve:

1. Promote exact five-group labels for Challenge 4 before another runtime tuning pass.
2. Build a true serpentine primitive with documented arc width, drop depth, and exit side rather than parameter-tweaking the current path.
3. Add high-bonus readability scoring so serpentine complexity stays playable.
4. Compare Aurora object tracks against target-video object tracks at group level, not only stage average.
5. Keep this out of runtime promotion until target-video and human-visible scores both improve.

Challenge 5 / Internal Stage Marker 19

Current score: 4.4/10

Target state:

• This should distinguish itself with pink/green cascade motion, alternating group identity, and a more memorable lower-field pass.

Current Aurora state:

• Movement and graphics are slightly higher than the average, but target contract and target-video fit are poor.
• Earlier candidate evidence suggested human-perfect potential could improve, but human-visible readability got worse.

Differences and conformance gap:

• This is the clearest example of the "wrong metric can improve" problem: scoreability can rise while visible conformance falls.
• The current target-video fit of 2.7/10 is too low for gameplay promotion.

Most specific work to improve:

1. Separate scoreability optimization from conformance promotion in the sweep output.
2. Add exact cascade group labels from the target video: group order, alternating family, entry side, exit side, and lower-field band.
3. Add spacing-aware candidate generation so improvements do not become clumped visual noise.
4. Add specialty alien crop and motion targets for the pink/green family.
5. Do not tune runtime until the scorer can explain why target-video fit is so low.

Challenge 6 / Internal Stage Marker 23

Current score: 4.2/10

Target state:

• This should emphasize green ladder rhythm and split exits with staggered group timing that a player can read.

Current Aurora state:

• The stage lands on the expected broad reference family, but target contract fit and target-video fit are still low.
• Human-perfect and human-visible sweep scoring are not yet available for the late stages, making this hard to promote safely.

Differences and conformance gap:

• The broad family is present, but ladder rung timing, split exits, and group separation are not precise enough.
• Late-stage evidence is less complete than the first five.

Most specific work to improve:

1. Add human-perfect and human-visible probes for Challenge 6.
2. Label target ladder rungs and split-exit sides at group level.
3. Add a reusable ladder/split primitive for future Aurora variants and other games.
4. Produce synchronized target/current video for the full challenge window.
5. Only then run candidate sweeps with a promotion gate that includes staggered spacing and split-route readability.

Challenge 7 / Internal Stage Marker 27

Current score: 4.3/10

Target state:

• This should be a yellow diagonal fan with an obvious diagonal scoring lane and a distinct late-stage identity.

Current Aurora state:

• Movement and graphics are moderate, but the target contract fit is weak.
• The best reference is a later group in the right family, which suggests partial identity but not full group-order fidelity.

Differences and conformance gap:

• The target shape is only partially recognized; current scoring opportunities are not clearly aligned to the diagonal fan.
• The shot-opportunity score is the lowest of the eight at 5.1/10.

Most specific work to improve:

1. Add a diagonal-lane shot-opportunity probe instead of treating center-lane waiting as sufficient.
2. Label all five target groups and require group-order match, not just broad family match.
3. Add a diagonal fan primitive with explicit lane angle, group spacing, and exit timing.
4. Add human-visible scoring for late stages before accepting runtime changes.
5. Raise target contract fit before optimizing local gameplay feel.

Challenge 8 / Internal Stage Marker 31

Current score: 4.4/10

Target state:

• This should act as a compact late-loop capstone with blue/purple identity, stronger spectacle, and no return to early-stage vocabulary.

Current Aurora state:

• The runtime has a named blue/purple finale contract, but the target contract fit is low.
• Candidate search found expected-score lifts, but target-video fit could regress, so the current scorer is not safe enough for promotion.

Differences and conformance gap:

• The broad late-loop idea exists, but target-video fit is only 3.1/10.
• This stage needs both better target labels and a stronger path-shape candidate strategy.

Most specific work to improve:

1. Resolve the expected-score versus target-video-fit conflict in the candidate scorer.
2. Add direct target object labels for all five groups before another runtime sweep.
3. Broaden candidate strategy with new path-shape constants instead of only arc/drop retuning.
4. Add active blue/purple specialty sprite-motion evidence.
5. Keep this as a later-stage capstone work item until the first five challenges have visible wins.

Recommended Order Of Work

1. Regenerate and preserve challenge-stage conformance artifacts after every runtime challenge change.
2. Fix public naming to distinguish regular stages from challenge windows while preserving internal markers.
3. Promote Challenge 1 five-group labels and synchronized target/current video.
4. Use the improved Challenge 2 identity as the first near-term gameplay target: tighten cross-sweep path precision and active graphics.
5. Promote Challenge 3 boss-led active-motion labels and score them as animation, not labels.
6. Add human-perfect and human-visible gates to Challenges 6-8 before tuning those late stages.
7. Split candidate scores into player-scoreability, target-video fit, and human-visible readability so the optimizer cannot hide regressions.
8. Add spacing/bunching and magic-appearance penalties into strict conformance.
9. Build reusable path primitives: peel, cross-sweep, boss-led hook, serpentine, cascade, ladder/split, diagonal fan, and finale loop.
10. Use the first five challenges as the quality proving ground; generalize the movement grammar to later challenges only after at least one early challenge reaches a visible 5.0+ strict score without safety regression.

Success Criteria For The Next Major Lift

• Raise strict challenge-stage conformance from 4.3/10 to at least 5.0/10 by making one early challenge visibly target-like.
• Raise target-video object-track fit from 3.6/10 to at least 4.5/10 without lowering safety.
• Raise alien/stage novelty from 3.9/10 by scoring actual active sprite motion and specialty alien phases.
• Preserve 10/10 safety: no enemy shots, no attack starts, no ship losses.
• Produce target/current synchronized clips and contact sheets for every promoted gameplay change.
• Promote no candidate unless it passes target-video, human-visible, human-perfect, spacing, and identity gates together.

Generated: 2026-05-19T13:35:43.792Z Commit: 97ab4605 Branch: codex/macbook-conformance-investment-review

Summary

The generated index now contains 39 ordered rows: 31 regular levels and 8 challenging stages. Every row has a current Aurora runtime screenshot, an actual Galaga target-gameplay frame, and paired 10-second current/target gameplay clips for inline motion review. The important caveat is grounding quality: 14/39 target rows are exact ingested windows, while 25 regular-level rows use representative actual Galaga gameplay until a full per-level normal-stage corpus is ingested. Challenge-stage visual conformance remains the most human-visible failure: exact challenge target screenshots exist, but the strict challenge score is only 4.2/10.

Level 1

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-01.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-01.webm

Target source: Galaga Stage 1 opening rack entry; exact target row.

Current roles: Bee / Zako: 20, Butterfly / escort: 16, Boss Galaga: 4

Conformance read: This row has an exact ingested Galaga regular-stage target frame. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band opening; target family regular-stage-entry.

Current read: Bee / Zako: 20, Butterfly / escort: 16, Boss Galaga: 4

Critical gap: Exact frame exists, but the current index still uses a single mid-level snapshot; next precision requires temporal path overlays and per-entry group labels.

Next: Add target/current motion overlays for the first two entry groups and boss/capture windows.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png

Level 2

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-02.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-02.webm

Target source: Galaga Stage 2 early formation entry; exact target row.

Current roles: Bee / Zako: 20, Butterfly / escort: 16, Boss Galaga: 4

Conformance read: This row has an exact ingested Galaga regular-stage target frame. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band opening; target family regular-stage-entry.

Current read: Bee / Zako: 20, Butterfly / escort: 16, Boss Galaga: 4

Critical gap: Exact frame exists, but the current index still uses a single mid-level snapshot; next precision requires temporal path overlays and per-entry group labels.

Next: Add target/current motion overlays for the first two entry groups and boss/capture windows.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png

Level 3

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-03.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-03.webm

Target source: Representative Galaga target: Stage 4 post-challenge rack entry; representative actual target gameplay row.

Current roles: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 4

Conformance read: This row is honest about target debt: it uses actual Galaga gameplay from Representative Galaga target: Stage 4 post-challenge rack entry, but not an exact per-level target capture. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band early-pressure; target family regular-stage-entry.

Current read: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 4

Critical gap: Exact target frame for Level 3 is missing from the ingested corpus; this should not be scored as fully grounded visual conformance.

Next: Ingest or extract a precise Galaga Level 3 window, then replace this representative target row.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Challenging Stage 3-4

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/challenge-01.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/challenge-01.webm

Target source: Galaga Challenging Stage 3-4; exact target row.

Current roles: Bee / Zako: 20, Butterfly / escort: 20

Conformance read: This is a no-combat bonus exhibition, so safety is only the floor. Current Aurora still reads weak on movement grammar, alien novelty, and graphical spectacle unless the strict scores here are materially higher than 6/10.

Variation read: Introductory challenge teaches vertical columns, simple side hooks, upper-band score windows, and the perfect-result loop.

Current read: first-challenge-peel / first-challenge-peel; lanes bee, bee, bee, bee, but, but, but, but; first-wave types bee, bee, bee, bee, but, but, but, but; visual families classic; strict movement 4.2/10, graphics 4.3/10, alien novelty 3.9/10, shot opportunity 5.8/10, target-contract fit 6.8/10.

Critical gap: Alien/stage novelty is only 3.9/10: the current type mix does not yet prove Galaga-like new alien introductions, memorable challenge-specific roles, or distinct bonus-stage learning patterns.

Next: Protect the first-challenge bee/butterfly line contract, then tune path length, turn count, and rack-slot precision against challenge-1 arrival and late-wave labels.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png

Level 4

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-04.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-04.webm

Target source: Galaga Stage 4 post-challenge rack entry; exact target row.

Current roles: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 4

Conformance read: This row has an exact ingested Galaga regular-stage target frame. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band early-pressure; target family regular-stage-entry.

Current read: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 4

Critical gap: Exact frame exists, but the current index still uses a single mid-level snapshot; next precision requires temporal path overlays and per-entry group labels.

Next: Add target/current motion overlays for the first two entry groups and boss/capture windows.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Level 5

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-05.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-05.webm

Target source: Galaga Stage 5 opening rack entry; exact target row.

Current roles: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 4

Conformance read: This row has an exact ingested Galaga regular-stage target frame. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band early-pressure; target family regular-stage-entry.

Current read: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 4

Critical gap: Exact frame exists, but the current index still uses a single mid-level snapshot; next precision requires temporal path overlays and per-entry group labels.

Next: Add target/current motion overlays for the first two entry groups and boss/capture windows.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Level 6

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-06.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-06.webm

Target source: Galaga Stage 6 capture and boss pressure; exact target row.

Current roles: Specialty challenge alien: 20, Butterfly / escort: 15, Boss Galaga: 4

Conformance read: This row has an exact ingested Galaga regular-stage target frame. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band early-pressure; target family regular-stage-entry.

Current read: Specialty challenge alien: 20, Butterfly / escort: 15, Boss Galaga: 4

Critical gap: Exact frame exists, but the current index still uses a single mid-level snapshot; next precision requires temporal path overlays and per-entry group labels.

Next: Add target/current motion overlays for the first two entry groups and boss/capture windows.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Level 7

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-07.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-07.webm

Target source: Representative Galaga target: Stage 6 capture and boss pressure; representative actual target gameplay row.

Current roles: Specialty challenge alien: 20, Butterfly / escort: 15, Boss Galaga: 4

Conformance read: This row is honest about target debt: it uses actual Galaga gameplay from Representative Galaga target: Stage 6 capture and boss pressure, but not an exact per-level target capture. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band capture-pressure; target family regular-stage-entry.

Current read: Specialty challenge alien: 20, Butterfly / escort: 15, Boss Galaga: 4

Critical gap: Exact target frame for Level 7 is missing from the ingested corpus; this should not be scored as fully grounded visual conformance.

Next: Ingest or extract a precise Galaga Level 7 window, then replace this representative target row.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Challenging Stage 7-8

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/challenge-02.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/challenge-02.webm

Target source: Galaga Challenging Stage 7-8; exact target row.

Current roles: Boss Galaga: 10, Butterfly / escort: 10, Captured / rogue fighter: 10, Specialty challenge alien: 8, Bee / Zako: 2

Conformance read: This is a no-combat bonus exhibition, so safety is only the floor. Current Aurora still reads weak on movement grammar, alien novelty, and graphical spectacle unless the strict scores here are materially higher than 6/10.

Variation read: Adds side-crossing and vertical red-column trains; groups arrive through visible route commitments rather than appearing in place.

Current read: scorpion-cross-sweep / cross-sweep; lanes but, boss, rogue, bee, bee, rogue, boss, but; first-wave types but, boss, rogue, bee, bee, rogue, boss, but; visual families classic; strict movement 4.3/10, graphics 4/10, alien novelty 3.9/10, shot opportunity 5.5/10, target-contract fit 6.3/10.

Critical gap: Alien/stage novelty is only 3.9/10: the current type mix does not yet prove Galaga-like new alien introductions, memorable challenge-specific roles, or distinct bonus-stage learning patterns.

Next: Tune challenge 2 toward the denser mixed-novelty-line reference instead of relying on a generic cross-sweep.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Captured / rogue fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/rogue-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png

Level 8

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-08.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-08.webm

Target source: Representative Galaga target: Stage 10 opening rack entry; representative actual target gameplay row.

Current roles: Specialty challenge alien: 20, Butterfly / escort: 15, Boss Galaga: 4

Conformance read: This row is honest about target debt: it uses actual Galaga gameplay from Representative Galaga target: Stage 10 opening rack entry, but not an exact per-level target capture. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band capture-pressure; target family regular-stage-entry.

Current read: Specialty challenge alien: 20, Butterfly / escort: 15, Boss Galaga: 4

Critical gap: Exact target frame for Level 8 is missing from the ingested corpus; this should not be scored as fully grounded visual conformance.

Next: Ingest or extract a precise Galaga Level 8 window, then replace this representative target row.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Level 9

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-09.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-09.webm

Target source: Representative Galaga target: Stage 10 opening rack entry; representative actual target gameplay row.

Current roles: Specialty challenge alien: 20, Butterfly / escort: 15, Boss Galaga: 4

Conformance read: This row is honest about target debt: it uses actual Galaga gameplay from Representative Galaga target: Stage 10 opening rack entry, but not an exact per-level target capture. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band capture-pressure; target family regular-stage-entry.

Current read: Specialty challenge alien: 20, Butterfly / escort: 15, Boss Galaga: 4

Critical gap: Exact target frame for Level 9 is missing from the ingested corpus; this should not be scored as fully grounded visual conformance.

Next: Ingest or extract a precise Galaga Level 9 window, then replace this representative target row.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Level 10

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-10.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-10.webm

Target source: Galaga Stage 10 opening rack entry; exact target row.

Current roles: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 4

Conformance read: This row has an exact ingested Galaga regular-stage target frame. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band capture-pressure; target family regular-stage-entry.

Current read: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 4

Critical gap: Exact frame exists, but the current index still uses a single mid-level snapshot; next precision requires temporal path overlays and per-entry group labels.

Next: Add target/current motion overlays for the first two entry groups and boss/capture windows.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Level 11

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-11.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-11.webm

Target source: Representative Galaga target: Stage 10 opening rack entry; representative actual target gameplay row.

Current roles: Specialty challenge alien: 20, Butterfly / escort: 15, Boss Galaga: 4

Conformance read: This row is honest about target debt: it uses actual Galaga gameplay from Representative Galaga target: Stage 10 opening rack entry, but not an exact per-level target capture. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band mid-loop; target family regular-stage-entry.

Current read: Specialty challenge alien: 20, Butterfly / escort: 15, Boss Galaga: 4

Critical gap: Exact target frame for Level 11 is missing from the ingested corpus; this should not be scored as fully grounded visual conformance.

Next: Ingest or extract a precise Galaga Level 11 window, then replace this representative target row.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Challenging Stage 11-12

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/challenge-03.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/challenge-03.webm

Target source: Galaga Challenging Stage 11-12; exact target row.

Current roles: Boss Galaga: 12, Butterfly / escort: 9, Captured / rogue fighter: 9, Challenge dragonfly family: 4, Specialty challenge alien: 2

Conformance read: This is a no-combat bonus exhibition, so safety is only the floor. Current Aurora still reads weak on movement grammar, alien novelty, and graphical spectacle unless the strict scores here are materially higher than 6/10.

Variation read: Blue side hooks and green novelty shapes make the third challenge visually distinct and less column-like.

Current read: stingray-crown-hook-hybrid / hook-arc; lanes boss, rogue, but, bee, boss, bee, but, rogue; first-wave types boss, rogue, but, bee, boss, bee, but, rogue; visual families dragonfly; strict movement 4.3/10, graphics 4.5/10, alien novelty 3.9/10, shot opportunity 5.6/10, target-contract fit 7/10.

Critical gap: Alien/stage novelty is only 3.9/10: the current type mix does not yet prove Galaga-like new alien introductions, memorable challenge-specific roles, or distinct bonus-stage learning patterns.

Next: Promote challenge-3 boss-led reference phases and score dragonfly visual novelty as animation, not only family label.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Challenge dragonfly family: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png
• Captured / rogue fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/rogue-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png

Level 12

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-12.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-12.webm

Target source: Representative Galaga target: Stage 10 opening rack entry; representative actual target gameplay row.

Current roles: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 3

Conformance read: This row is honest about target debt: it uses actual Galaga gameplay from Representative Galaga target: Stage 10 opening rack entry, but not an exact per-level target capture. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band mid-loop; target family regular-stage-entry.

Current read: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 3

Critical gap: Exact target frame for Level 12 is missing from the ingested corpus; this should not be scored as fully grounded visual conformance.

Next: Ingest or extract a precise Galaga Level 12 window, then replace this representative target row.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Level 13

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-13.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-13.webm

Target source: Representative Galaga target: Stage 10 opening rack entry; representative actual target gameplay row.

Current roles: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 3

Conformance read: This row is honest about target debt: it uses actual Galaga gameplay from Representative Galaga target: Stage 10 opening rack entry, but not an exact per-level target capture. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band mid-loop; target family regular-stage-entry.

Current read: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 3

Critical gap: Exact target frame for Level 13 is missing from the ingested corpus; this should not be scored as fully grounded visual conformance.

Next: Ingest or extract a precise Galaga Level 13 window, then replace this representative target row.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Level 14

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-14.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-14.webm

Target source: Representative Galaga target: Stage 10 opening rack entry; representative actual target gameplay row.

Current roles: Specialty challenge alien: 20, Butterfly / escort: 15, Boss Galaga: 4

Conformance read: This row is honest about target debt: it uses actual Galaga gameplay from Representative Galaga target: Stage 10 opening rack entry, but not an exact per-level target capture. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band mid-loop; target family regular-stage-entry.

Current read: Specialty challenge alien: 20, Butterfly / escort: 15, Boss Galaga: 4

Critical gap: Exact target frame for Level 14 is missing from the ingested corpus; this should not be scored as fully grounded visual conformance.

Next: Ingest or extract a precise Galaga Level 14 window, then replace this representative target row.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Level 15

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-15.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-15.webm

Target source: Representative Galaga target: Stage 10 opening rack entry; representative actual target gameplay row.

Current roles: Specialty challenge alien: 19, Butterfly / escort: 16, Boss Galaga: 4

Conformance read: This row is honest about target debt: it uses actual Galaga gameplay from Representative Galaga target: Stage 10 opening rack entry, but not an exact per-level target capture. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band mid-loop; target family regular-stage-entry.

Current read: Specialty challenge alien: 19, Butterfly / escort: 16, Boss Galaga: 4

Critical gap: Exact target frame for Level 15 is missing from the ingested corpus; this should not be scored as fully grounded visual conformance.

Next: Ingest or extract a precise Galaga Level 15 window, then replace this representative target row.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Challenging Stage 15-16

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/challenge-04.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/challenge-04.webm

Target source: Galaga Challenging Stage 15-16; exact target row.

Current roles: Butterfly / escort: 7, Boss Galaga: 6, Captured / rogue fighter: 6, Challenge dragonfly family: 4, Specialty challenge alien: 4

Conformance read: This is a no-combat bonus exhibition, so safety is only the floor. Current Aurora still reads weak on movement grammar, alien novelty, and graphical spectacle unless the strict scores here are materially higher than 6/10.

Variation read: Late challenge starts showing long pink serpentine arcs and separate green specialty entries.

Current read: pink-serpentine-late / pink-serpentine; lanes boss, rogue, but, bee, bee, but, rogue, boss; first-wave types boss, rogue, but, bee, bee, but, rogue, boss; visual families galboss; strict movement 4.3/10, graphics 4.5/10, alien novelty 3.9/10, shot opportunity 5.5/10, target-contract fit 4.5/10.

Critical gap: Alien/stage novelty is only 3.9/10: the current type mix does not yet prove Galaga-like new alien introductions, memorable challenge-specific roles, or distinct bonus-stage learning patterns.

Next: Promote the Challenge 4 pink-serpentine window into five group labels and tune the runtime path so all groups keep a readable serpentine score lane.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png
• Captured / rogue fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/rogue-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Challenge dragonfly family: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Level 16

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-16.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-16.webm

Target source: Representative Galaga target: Stage 10 opening rack entry; representative actual target gameplay row.

Current roles: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 4

Conformance read: This row is honest about target debt: it uses actual Galaga gameplay from Representative Galaga target: Stage 10 opening rack entry, but not an exact per-level target capture. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band mid-loop; target family regular-stage-entry.

Current read: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 4

Critical gap: Exact target frame for Level 16 is missing from the ingested corpus; this should not be scored as fully grounded visual conformance.

Next: Ingest or extract a precise Galaga Level 16 window, then replace this representative target row.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Level 17

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-17.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-17.webm

Target source: Representative Galaga target: Stage 10 opening rack entry; representative actual target gameplay row.

Current roles: Specialty challenge alien: 20, Butterfly / escort: 15, Boss Galaga: 4

Conformance read: This row is honest about target debt: it uses actual Galaga gameplay from Representative Galaga target: Stage 10 opening rack entry, but not an exact per-level target capture. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band mid-loop; target family regular-stage-entry.

Current read: Specialty challenge alien: 20, Butterfly / escort: 15, Boss Galaga: 4

Critical gap: Exact target frame for Level 17 is missing from the ingested corpus; this should not be scored as fully grounded visual conformance.

Next: Ingest or extract a precise Galaga Level 17 window, then replace this representative target row.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Level 18

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-18.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-18.webm

Target source: Representative Galaga target: Stage 10 opening rack entry; representative actual target gameplay row.

Current roles: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 4

Conformance read: This row is honest about target debt: it uses actual Galaga gameplay from Representative Galaga target: Stage 10 opening rack entry, but not an exact per-level target capture. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band mid-loop; target family regular-stage-entry.

Current read: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 4

Critical gap: Exact target frame for Level 18 is missing from the ingested corpus; this should not be scored as fully grounded visual conformance.

Next: Ingest or extract a precise Galaga Level 18 window, then replace this representative target row.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Level 19

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-19.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-19.webm

Target source: Representative Galaga target: Stage 10 opening rack entry; representative actual target gameplay row.

Current roles: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 3

Conformance read: This row is honest about target debt: it uses actual Galaga gameplay from Representative Galaga target: Stage 10 opening rack entry, but not an exact per-level target capture. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band late-loop; target family regular-stage-entry.

Current read: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 3

Critical gap: Exact target frame for Level 19 is missing from the ingested corpus; this should not be scored as fully grounded visual conformance.

Next: Ingest or extract a precise Galaga Level 19 window, then replace this representative target row.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Challenging Stage 19-20

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/challenge-05.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/challenge-05.webm

Target source: Galaga Challenging Stage 19-20; exact target row.

Current roles: Butterfly / escort: 7, Captured / rogue fighter: 4, Boss Galaga: 3, Specialty challenge alien: 2, Challenge dragonfly family: 1

Conformance read: This is a no-combat bonus exhibition, so safety is only the floor. Current Aurora still reads weak on movement grammar, alien novelty, and graphical spectacle unless the strict scores here are materially higher than 6/10.

Variation read: Continues specialty novelty with pink arcs, green lower-field entries, and a clearer split between group identities.

Current read: pink-green-cascade / pink-green-cascade; lanes rogue, boss, but, bee, boss, but, bee, rogue; first-wave types rogue, boss, but, bee, boss, but, bee, rogue; visual families galboss; strict movement 4.6/10, graphics 4.5/10, alien novelty 3.9/10, shot opportunity 5.5/10, target-contract fit 4.3/10.

Critical gap: Alien/stage novelty is only 3.9/10: the current type mix does not yet prove Galaga-like new alien introductions, memorable challenge-specific roles, or distinct bonus-stage learning patterns.

Next: Promote the Challenge 5 pink/green cascade window into five group labels and tune lower-field pass timing against those labels.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png
• Captured / rogue fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/rogue-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Challenge dragonfly family: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Level 20

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-20.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-20.webm

Target source: Representative Galaga target: Stage 10 opening rack entry; representative actual target gameplay row.

Current roles: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 4

Conformance read: This row is honest about target debt: it uses actual Galaga gameplay from Representative Galaga target: Stage 10 opening rack entry, but not an exact per-level target capture. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band late-loop; target family regular-stage-entry.

Current read: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 4

Critical gap: Exact target frame for Level 20 is missing from the ingested corpus; this should not be scored as fully grounded visual conformance.

Next: Ingest or extract a precise Galaga Level 20 window, then replace this representative target row.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Level 21

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-21.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-21.webm

Target source: Representative Galaga target: Stage 10 opening rack entry; representative actual target gameplay row.

Current roles: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 4

Conformance read: This row is honest about target debt: it uses actual Galaga gameplay from Representative Galaga target: Stage 10 opening rack entry, but not an exact per-level target capture. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band late-loop; target family regular-stage-entry.

Current read: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 4

Critical gap: Exact target frame for Level 21 is missing from the ingested corpus; this should not be scored as fully grounded visual conformance.

Next: Ingest or extract a precise Galaga Level 21 window, then replace this representative target row.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Level 22

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-22.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-22.webm

Target source: Representative Galaga target: Stage 10 opening rack entry; representative actual target gameplay row.

Current roles: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 4

Conformance read: This row is honest about target debt: it uses actual Galaga gameplay from Representative Galaga target: Stage 10 opening rack entry, but not an exact per-level target capture. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band late-loop; target family regular-stage-entry.

Current read: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 4

Critical gap: Exact target frame for Level 22 is missing from the ingested corpus; this should not be scored as fully grounded visual conformance.

Next: Ingest or extract a precise Galaga Level 22 window, then replace this representative target row.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Level 23

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-23.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-23.webm

Target source: Representative Galaga target: Stage 10 opening rack entry; representative actual target gameplay row.

Current roles: Specialty challenge alien: 20, Butterfly / escort: 15, Boss Galaga: 4

Conformance read: This row is honest about target debt: it uses actual Galaga gameplay from Representative Galaga target: Stage 10 opening rack entry, but not an exact per-level target capture. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band late-loop; target family regular-stage-entry.

Current read: Specialty challenge alien: 20, Butterfly / escort: 15, Boss Galaga: 4

Critical gap: Exact target frame for Level 23 is missing from the ingested corpus; this should not be scored as fully grounded visual conformance.

Next: Ingest or extract a precise Galaga Level 23 window, then replace this representative target row.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Challenging Stage 23-24

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/challenge-06.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/challenge-06.webm

Target source: Galaga Challenging Stage 23-24; exact target row.

Current roles: Captured / rogue fighter: 9, Challenge dragonfly family: 7, Boss Galaga: 6, Butterfly / escort: 5, Challenge mosquito family: 2

Conformance read: This is a no-combat bonus exhibition, so safety is only the floor. Current Aurora still reads weak on movement grammar, alien novelty, and graphical spectacle unless the strict scores here are materially higher than 6/10.

Variation read: Green ladder and split entries emphasize staggered group timing and route separation.

Current read: green-ladder-split / green-ladder-split; lanes bee, but, rogue, boss, boss, rogue, but, bee; first-wave types bee, but, rogue, boss, boss, rogue, but, bee; visual families dragonfly; strict movement 4.3/10, graphics 4.4/10, alien novelty 3.9/10, shot opportunity 5.9/10, target-contract fit 4.6/10.

Critical gap: Alien/stage novelty is only 3.9/10: the current type mix does not yet prove Galaga-like new alien introductions, memorable challenge-specific roles, or distinct bonus-stage learning patterns.

Next: Rebuild Challenge 6 around green ladder and split-exit timing until its measured vector hits challenge-6-green-ladder-split-group-1.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Captured / rogue fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/rogue-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Challenge dragonfly family: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png
• Challenge mosquito family: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-mosquito.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Level 24

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-24.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-24.webm

Target source: Representative Galaga target: Stage 10 opening rack entry; representative actual target gameplay row.

Current roles: Specialty challenge alien: 19, Butterfly / escort: 16, Boss Galaga: 4

Conformance read: This row is honest about target debt: it uses actual Galaga gameplay from Representative Galaga target: Stage 10 opening rack entry, but not an exact per-level target capture. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band late-loop; target family regular-stage-entry.

Current read: Specialty challenge alien: 19, Butterfly / escort: 16, Boss Galaga: 4

Critical gap: Exact target frame for Level 24 is missing from the ingested corpus; this should not be scored as fully grounded visual conformance.

Next: Ingest or extract a precise Galaga Level 24 window, then replace this representative target row.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Level 25

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-25.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-25.webm

Target source: Representative Galaga target: Stage 10 opening rack entry; representative actual target gameplay row.

Current roles: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 4

Conformance read: This row is honest about target debt: it uses actual Galaga gameplay from Representative Galaga target: Stage 10 opening rack entry, but not an exact per-level target capture. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band late-loop; target family regular-stage-entry.

Current read: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 4

Critical gap: Exact target frame for Level 25 is missing from the ingested corpus; this should not be scored as fully grounded visual conformance.

Next: Ingest or extract a precise Galaga Level 25 window, then replace this representative target row.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Level 26

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-26.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-26.webm

Target source: Representative Galaga target: Stage 10 opening rack entry; representative actual target gameplay row.

Current roles: Specialty challenge alien: 20, Butterfly / escort: 15, Boss Galaga: 4

Conformance read: This row is honest about target debt: it uses actual Galaga gameplay from Representative Galaga target: Stage 10 opening rack entry, but not an exact per-level target capture. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band late-loop; target family regular-stage-entry.

Current read: Specialty challenge alien: 20, Butterfly / escort: 15, Boss Galaga: 4

Critical gap: Exact target frame for Level 26 is missing from the ingested corpus; this should not be scored as fully grounded visual conformance.

Next: Ingest or extract a precise Galaga Level 26 window, then replace this representative target row.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Level 27

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-27.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-27.webm

Target source: Representative Galaga target: Stage 10 opening rack entry; representative actual target gameplay row.

Current roles: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 4

Conformance read: This row is honest about target debt: it uses actual Galaga gameplay from Representative Galaga target: Stage 10 opening rack entry, but not an exact per-level target capture. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band late-loop; target family regular-stage-entry.

Current read: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 4

Critical gap: Exact target frame for Level 27 is missing from the ingested corpus; this should not be scored as fully grounded visual conformance.

Next: Ingest or extract a precise Galaga Level 27 window, then replace this representative target row.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Challenging Stage 27-28

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/challenge-07.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/challenge-07.webm

Target source: Galaga Challenging Stage 27-28; exact target row.

Current roles: No active enemy roles visible in this sampled current frame.

Conformance read: This is a no-combat bonus exhibition, so safety is only the floor. Current Aurora still reads weak on movement grammar, alien novelty, and graphical spectacle unless the strict scores here are materially higher than 6/10.

Variation read: Yellow diagonal trains create a memorable late-stage fan/diagonal score lane.

Current read: yellow-diagonal-fan / yellow-diagonal-fan; lanes boss, bee, but, rogue, rogue, but, bee, boss; first-wave types boss, bee, but, rogue, rogue, but, bee, boss; visual families crown; strict movement 4.2/10, graphics 4.3/10, alien novelty 3.9/10, shot opportunity 4.5/10, target-contract fit 4.7/10.

Critical gap: Alien/stage novelty is only 3.9/10: the current type mix does not yet prove Galaga-like new alien introductions, memorable challenge-specific roles, or distinct bonus-stage learning patterns.

Next: Rebuild Challenge 7 around the yellow diagonal fan so the best match lands on challenge-7-yellow-diagonal-fan-group-1, not the finale label.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Level 28

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-28.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-28.webm

Target source: Representative Galaga target: Stage 10 opening rack entry; representative actual target gameplay row.

Current roles: Specialty challenge alien: 20, Butterfly / escort: 15, Boss Galaga: 4

Conformance read: This row is honest about target debt: it uses actual Galaga gameplay from Representative Galaga target: Stage 10 opening rack entry, but not an exact per-level target capture. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band late-loop; target family regular-stage-entry.

Current read: Specialty challenge alien: 20, Butterfly / escort: 15, Boss Galaga: 4

Critical gap: Exact target frame for Level 28 is missing from the ingested corpus; this should not be scored as fully grounded visual conformance.

Next: Ingest or extract a precise Galaga Level 28 window, then replace this representative target row.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Level 29

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-29.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-29.webm

Target source: Representative Galaga target: Stage 10 opening rack entry; representative actual target gameplay row.

Current roles: Specialty challenge alien: 20, Butterfly / escort: 15, Boss Galaga: 4

Conformance read: This row is honest about target debt: it uses actual Galaga gameplay from Representative Galaga target: Stage 10 opening rack entry, but not an exact per-level target capture. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band late-loop; target family regular-stage-entry.

Current read: Specialty challenge alien: 20, Butterfly / escort: 15, Boss Galaga: 4

Critical gap: Exact target frame for Level 29 is missing from the ingested corpus; this should not be scored as fully grounded visual conformance.

Next: Ingest or extract a precise Galaga Level 29 window, then replace this representative target row.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Level 30

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-30.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-30.webm

Target source: Representative Galaga target: Stage 10 opening rack entry; representative actual target gameplay row.

Current roles: Specialty challenge alien: 20, Butterfly / escort: 15, Boss Galaga: 4

Conformance read: This row is honest about target debt: it uses actual Galaga gameplay from Representative Galaga target: Stage 10 opening rack entry, but not an exact per-level target capture. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band late-loop; target family regular-stage-entry.

Current read: Specialty challenge alien: 20, Butterfly / escort: 15, Boss Galaga: 4

Critical gap: Exact target frame for Level 30 is missing from the ingested corpus; this should not be scored as fully grounded visual conformance.

Next: Ingest or extract a precise Galaga Level 30 window, then replace this representative target row.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Level 31

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/level-31.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/level-31.webm

Target source: Representative Galaga target: Stage 10 opening rack entry; representative actual target gameplay row.

Current roles: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 3

Conformance read: This row is honest about target debt: it uses actual Galaga gameplay from Representative Galaga target: Stage 10 opening rack entry, but not an exact per-level target capture. The useful review question is whether Aurora’s current frame has a readable rack/entry state, recognizable bee/butterfly/boss roles, and stage-band pressure rather than only a different background.

Variation read: Regular level band late-loop; target family regular-stage-entry.

Current read: Specialty challenge alien: 20, Butterfly / escort: 16, Boss Galaga: 3

Critical gap: Exact target frame for Level 31 is missing from the ingested corpus; this should not be scored as fully grounded visual conformance.

Next: Ingest or extract a precise Galaga Level 31 window, then replace this representative target row.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Bee / Zako: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/bee-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-zako-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png

Challenging Stage 31-32

Private companion store only; not exposed from the public repo.

Private companion store only; not exposed from the public repo.

Aurora 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-current-videos/challenge-08.webm

Galaga target 10s video: reference-artifacts/analyses/level-visual-conformance-index/latest-target-videos/challenge-08.webm

Target source: Galaga Challenging Stage 31-32; exact target row.

Current roles: Butterfly / escort: 5, Captured / rogue fighter: 4, Specialty challenge alien: 4, Boss Galaga: 3

Conformance read: This is a no-combat bonus exhibition, so safety is only the floor. Current Aurora still reads weak on movement grammar, alien novelty, and graphical spectacle unless the strict scores here are materially higher than 6/10.

Variation read: Final visible challenge window emphasizes blue/purple clustered arcs and a compact all-perfect finish.

Current read: blue-purple-finale / blue-purple-finale; lanes rogue, boss, bee, but, but, bee, boss, rogue; first-wave types rogue, boss, bee, but, but, bee, boss, rogue; visual families stingray; strict movement 4.4/10, graphics 4.5/10, alien novelty 3.9/10, shot opportunity 5.6/10, target-contract fit 4.3/10.

Critical gap: Alien/stage novelty is only 3.9/10: the current type mix does not yet prove Galaga-like new alien introductions, memorable challenge-specific roles, or distinct bonus-stage learning patterns.

Next: Refine the Challenge 8 blue/purple finale with fuller path length and compact late-loop timing.

Sprite / alien bitmap references:

• Player fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/player-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png
• Boss Galaga: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/boss-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-command-boss.png
• Specialty challenge alien: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/challenge-dragonfly.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-specialty-dive.png
• Butterfly / escort: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/but-line.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-butterfly-escort.png
• Captured / rogue fighter: current reference-artifacts/analyses/aurora-runtime-sprite-conformance/latest-crops/rogue-fighter.png; target reference-artifacts/analyses/galaga-reference-sprites/pixel-targets-0.1/galaga-player-fighter.png

Generated: 2026-05-18T20:10:36.556Z

This analysis turns the two user-supplied Galaga challenge videos into durable derived evidence for Aurora challenging-stage conformance. The original source videos analyzed here still remain outside the repository. A newer stage- isolated Challenging Stages.mp4 compilation is now preserved in-repo at reference-artifacts/preserved-sources/galaga-challenging-stages-isolated-2026-06-01/ as an additional raw source that complements this analysis when motion windows need to be cut from cleaner stage-by-stage footage. Committed artifacts here include contact sheets, timing windows, manual reads, and implementation contracts.

An additional repo-owned perfect-play challenge compilation now also exists at reference-artifacts/preserved-sources/galaga-perfect-challenging-stages-2026-06-01/. That source is best used for repeated perfect-result presentation and ideal-flow cadence review rather than highest-resolution motion extraction.

A newer and stronger repo-owned perfect-play compilation now also exists at reference-artifacts/preserved-sources/galaga-higher-quality-perfect-challenging-stages-2026-06-01/. That source should now be treated as the preferred repo-owned perfect-play lane for repeated perfect-result presentation and ideal-flow cadence review, while the earlier 320x240 perfect-play source remains a supplemental corroborating reference.

Summary

• Primary source: challenge-all2-single-ship-all-perfects
• Secondary source: challenging-stage-compilation
• Challenge windows: 8
• Late challenge windows now media-backed: 5/5
• Challenge-stage target readiness estimate: 4.5/10
• Status: late-challenge-media-windows-ready-for-labeling

What The Videos Show

• The challenge stages are non-combat score exhibitions: no enemy bullets, no ship-loss pressure, and repeated 40-hit perfect result screens.
• Each stage is built from visible group arrivals, not static appearances; the player reads routes and learns where the scoring lane will be.
• Progression adds new visual families and motion grammar: vertical columns, side hooks, crossing groups, serpentine arcs, green ladders, yellow diagonal fans, and blue/purple late clusters.
• The user-facing conformance gap in Aurora is mostly spectacle, route memory, and alien novelty, not the no-shot/no-kill safety rule.

Window Reads

Stage Improvement Plan

Challenge 1 / Stage Marker 3

Priority: high

Target family: classic-column-and-side-arc

Problem: Aurora early challenge stages have useful structure but need stronger five-group labels and clearer route contracts.

Strategy: Use as the teaching stage: five readable groups, no combat, clear upper-band score lanes, and a result cadence that lets the player understand 40/40.

Player impact: Early challenge stages should become clearer learning experiences with stronger perfect-bonus readability.

Challenge 2 / Stage Marker 7

Priority: high

Target family: red-column-blue-side-cross

Problem: Aurora early challenge stages have useful structure but need stronger five-group labels and clearer route contracts.

Strategy: Add distinct side-entry crossing groups and a red-column vertical train; preserve no-shot/no-collision behavior while increasing route memory.

Player impact: Early challenge stages should become clearer learning experiences with stronger perfect-bonus readability.

Challenge 3 / Stage Marker 11

Priority: high

Target family: blue-hook-green-novelty

Problem: Aurora early challenge stages have useful structure but need stronger five-group labels and clearer route contracts.

Strategy: Introduce a clearly new visual family and hooked side trajectories; score should reward anticipation of the hook, not pure center firing.

Player impact: Early challenge stages should become clearer learning experiences with stronger perfect-bonus readability.

Challenge 4 / Stage Marker 15

Priority: highest

Target family: pink-serpentine-and-green-entry

Problem: Aurora late challenge stages are currently too repetitive and are not media-backed by distinct late-stage Galaga choreography.

Strategy: Replace repeated boss-led loops with a long serpentine specialty arc; this is the first high-priority late-stage rebuild target.

Player impact: Late challenge stages should start feeling like memorable Galaga score exhibitions instead of repeated waves.

Challenge 5 / Stage Marker 19

Priority: highest

Target family: pink-green-cascade

Problem: Aurora late challenge stages are currently too repetitive and are not media-backed by distinct late-stage Galaga choreography.

Strategy: Build as a cascade stage: alternating specialty groups, lower-field pass risk, and stronger color/family novelty than Challenge 4.

Player impact: Late challenge stages should start feeling like memorable Galaga score exhibitions instead of repeated waves.

Challenge 6 / Stage Marker 23

Priority: highest

Target family: green-ladder-and-split

Problem: Aurora late challenge stages are currently too repetitive and are not media-backed by distinct late-stage Galaga choreography.

Strategy: Use staggered ladders and split exits; measure group spacing and route separation so the stage feels authored, not dense noise.

Player impact: Late challenge stages should start feeling like memorable Galaga score exhibitions instead of repeated waves.

Challenge 7 / Stage Marker 27

Priority: highest

Target family: yellow-diagonal-fan

Problem: Aurora late challenge stages are currently too repetitive and are not media-backed by distinct late-stage Galaga choreography.

Strategy: Add a yellow diagonal/fan stage with long path length, readable aim bands, and a visually obvious new family.

Player impact: Late challenge stages should start feeling like memorable Galaga score exhibitions instead of repeated waves.

Challenge 8 / Stage Marker 31

Priority: highest

Target family: blue-purple-finale

Problem: Aurora late challenge stages are currently too repetitive and are not media-backed by distinct late-stage Galaga choreography.

Strategy: Use as the late-loop capstone: blue/purple clustered arcs, compact timing, and no repeated early-stage motion vocabulary.

Player impact: Late challenge stages should start feeling like memorable Galaga score exhibitions instead of repeated waves.

Next Best Steps

1. Promote challenge-all2 windows into accepted five-group reference labels for Challenges 1-8.
2. Rebuild Aurora Challenge 4 first with the pink-serpentine contract because it is the most visible current late-stage embarrassment.
3. Then rebuild Challenge 7 with the yellow-diagonal-fan contract because it creates a strong player-visible novelty jump.
4. Add runtime probes that verify each Aurora challenge has five groups, no shots, no ship losses, distinct path families, and expected visual-family novelty.
5. Only after labels exist, lift challenge-stage target readiness from partial media evidence to direct trajectory scoring.

Generated: 2026-05-18T17:54:31.819Z

This is the project-facing inventory of online and local artifacts that best illustrate Aurora's Galaga-like target. It records what is already ingested, what is only a candidate, and what still blocks stronger challenge-stage conformance work.

Summary

• Overall target-artifact ingestion coverage: 5.1/10
• Challenge-stage target readiness: 4.5/10
• Critical sources: 5
• Critical sources not fully ingested: 5
• Existing local evidence anchors: 44
• Missing local evidence anchors: 0

Interpretation

Aurora now has media-backed windows for all tracked Galaga challenge stages, including Challenges 4-8. The bottleneck has moved from source acquisition to precision: each window needs five-group frame labels for entry side, path family, scoreable band, exit side, alien family, and perfect-bonus opportunity before direct trajectory scoring should rise.

Source Coverage

Challenge-Stage Priority Windows