Development lane Project Page Steven Woods

Aurora Galactica on Platinum

Aurora Galactica is the first shipped game on the Platinum browser-arcade platform and the active proof of a broader conformance program: reference ingestion, local harnessing, score dashboards, release lanes, and Codex-assisted development working together to make inspired arcade software more measurable and repeatable.

Release Version 1.4.1.1
Version label for the release location shown on this page.
Release Location Development lane
Generated from the development lane artifacts served locally and on hosted /dev.
Build Timestamp Jun 12 2026, 7:26 AM EDT
Timestamp of the build artifact for this release location, not necessarily the final public portfolio sync time.
Current Focus Aurora 1.4.1 is the production quality release for hosted `/production`, promoted from the accepted 1.4.1 beta review with Stage 3 keepers, audio/theme clarity, sign-in repair, and stronger release gates.
Current work on Aurora and Platinum pulled directly from the maintained release dashboard.
Latest Note 1.4.1 production quality release
Release note carried by this lane's generated artifacts.
Open this lane build Open beta build Open release dashboard Open conformance dashboard Open white paper Open overview deck Open project guide Open Platinum guide Open repository

Current state

Primary goal

Advance Aurora toward a stronger Galaga-like conformance gate while keeping Platinum clean as the reusable host and turning ingestion, harnesses, metrics, and resource economics into a repeatable path for future games.

How the platform helps

Platinum owns the shell, picker, hosted lane model, documentation surfaces, conformance dashboard, release framing, and shared harness substrate. Each game keeps its own rules, scoring, identity, and reference evidence.

Current conformance push

The highest-value work is audio/event feedback, level arc, alien entry and challenge-stage novelty, visual reference grounding, pressure replay precision, and Galaxy Guardians ingestion promotion before any broader second-game claim.

Release posture

Hosted /production now carries the live 1.4.0 public family, while hosted /dev and hosted /beta carry the next deliberate follow-through line as 1.4.0.1 and 1.4.0-beta.1. The next public release decision should come from that reviewed follow-through path, not from casual lane churn.

Aurora And Guardians Conformance

Use these lane-relative links when you want a direct conformance read for a specific game, or a quick compare entry point with both games visible at the top of the dashboard.

Aurora conformance

Open the hosted dashboard with Aurora selected immediately so the score, priorities, ingestion, and cost view all stay scoped to the shipped first cabinet.

Open Aurora conformance

Side-by-side starting point

Start from the shared dashboard when you want the top-level compare cards for both games first, then switch deeper into Aurora or Guardians from the selector.

Open compare view Open guide links

Short, medium, and long plan

The current recommendation is to keep the live 1.4.0 public family trustworthy, keep the documentation and dashboard story aligned, and use hosted /dev plus hosted /beta to shape the next deliberate follow-through bundle rather than blurring the lanes together. The plan below keeps release movement tied to measurable player-facing value and keeps the public project page aligned with the same release path as the software.

Short term: keep 1.4.0 trustworthy

Keep the shipped 1.4.0 public line stable, keep the conformance dashboard, documentation, application artifact scorecard, economics read, and release notes aligned, and use the current dev and beta lanes to review the next coherent improvement step.

Medium term: post-1.4.0 arcade depth

Shape the next beta family around alien entry variation, challenge-stage novelty, level arc, boss and formation grammar, visual reference grounding, audio and event feedback, and cleaner Platinum/application seams.

Long term: evidence to multi-game

Use 1.5.0 for shared video evidence and flight recorder, 1.6.0 for message-to-pilot cabinet polish, and 2.0 for a credible multi-game Platinum milestone built from ingestion-backed game-owned conformance packages.

Conformance system

This public page is meant to be read directly, not only used as a launcher. The conformance program compares Aurora against Galaga-family reference behavior through captured examples, derived metrics, repeatable harnesses, confidence notes, and resource accounting. The generated guide pages, dashboards, and source artifacts are drill-down surfaces behind the overview shown here.

Ingestion

Reference footage, source manifests, clipped windows, contact sheets, waveforms, event logs, alien indexes, audio cue indexes, stage summaries, and confidence notes become game-owned evidence rather than loose design memory.

Harnessing and scoring

Local CPU and browser harnesses measure movement, timing, audio, stage shape, boss formation grammar, visual surfaces, pressure windows, and shell behavior against reference-derived targets.

Resource economics

The project tracks local compute, browser work, artifact growth, and GPU-equivalent Codex/OpenAI/API usage so future conformance investment can be chosen by value, confidence, and cost.

Conformance readout

The score bars below are working measurements, not claims of perfection. A high number can still carry lower confidence when the scorer lacks enough reference windows or when the metric needs finer resolution. The goal is to make the gap visible, choose the next investment area, and then rerun the harnesses after each improvement.

Overall quality 8.8/10 Target >=9.3
Audio identity 7.3/10 Target >=7.5
Level arc 8.8/10 Target >=8.8
Boss entry and formation grammar 9.4/10 Target >=8.0 first gate; >=9.0 mature
Challenge-stage set-piece conformance: movement, graphics, alien novelty 4.3/10

The challenging stages should be spectacular safe Galaga-like bonus exhibitions. Current safety is good, but movement variation, alien novelty, and graphical conformance are not yet close.

Confidence: medium-high for gap direction; medium-low for exact lift estimate Target: >=5.0 before next beta claim; 6.0 after three authored challenges; 9.0+ mature
Audio identity, event feedback, and cue alignment 7.3/10

Largest current score gap and high user-experience impact: shots, explosions, boss damage, challenge results, capture/rescue feedback.

Confidence: medium-high Target: 7.5-8.0
Alien entry and broad challenge-stage novelty 8.2/10

Regular-stage alien entry, challenge-stage trajectories, and new-alien introduction still need stronger reference grounding; this broad metric should not mask the stricter challenge-stage gap.

Confidence: medium Target: 7.5 first gate; 9.0+ mature
Level arc and encounter shape 8.8/10

Controls whether long play feels like Galaga-like escalation rather than repeated pressure.

Confidence: medium-high Target: 8.8-9.0
Direct target sprite and impact feedback conformance 6/10

The player-visible ship, enemy, hit, and explosion shapes are a first-glance arcade quality signal. Static sprite proxy scores must not hide the stricter target-crop gap.

Confidence: medium Target: >=5.5 before next beta claim; >=7.5 mature preview
Boss entry and formation grammar 9.4/10

Boss entries, escorts, formation settling, and challenge set pieces are core Galaga choreography and directly affect whether stages feel authored.

Confidence: medium Target: 8.0-8.5 first gate; 9.0+ with path/slot extraction

Value versus compute

Conformance work is also an economics problem. Local CPU and browser runs should do as much of the repetitive measurement as possible, while Codex/OpenAI/model calls are treated as GPU-equivalent strategy, critique, harness-design, and interpretation work. The project tracks both so we can compare effort against measured score movement over time.

Measured runs 984 Harness and analysis executions in the current economics rollup.
Wall time 980.8 min Elapsed local time captured by conformance jobs.
CPU time 991.6 min Local compute used by the measurable assessment loop.
Artifacts 1442.7 MB Persisted evidence: traces, reports, frames, audio and charts.
Challenge-stage set-piece conformance: movement, graphics, alien novelty

Recent candidate sweeps improved measurement and search evidence, but no stage currently has a runtime keeper. Next: build richer movement primitives before runtime promotion; first target reference-spline-fit from 10 primitives, with first-five grammar at 25 group contracts and 25 reference-backed paths.

high; cpu, browser, gpu 162 runs; 24 min wall; 39.8 min CPU Value score: 8.40
Audio identity, event feedback, and cue alignment

Challenge Perfect runtime trial rejected perfect-clean-onset-soft-tail; do not directly promote the focused keeper. Do not promote Challenge Perfect from isolated onset/body candidates. Replace the next audio strategy with full-phrase/segment-boundary work: stabilize the scorer on canonical reference-vs-reference capture, then generate candidates that optimize onset, body, tail, and live capture segmentation together.

high; cpu, model-api, openai-api 317 runs; 254.3 min wall; 460.3 min CPU Value score: 2.67
Alien entry and broad challenge-stage novelty

Attack Regular-entry geometry separation: Minimum regular geometry distance 0.083; mean regular geometry distance 0.127; closest pair mid-run-entry-variant / late-run-cleanup-or-failure.

estimated; cpu, browser 113 runs; 26.4 min wall; 34.9 min CPU Value score: 0.00
Level arc and encounter shape

Use the top-ranked opportunity window to add or widen deterministic evidence before changing gameplay tuning.

low; cpu, browser 162 runs; 24 min wall; 39.8 min CPU Value score: 1.55
Direct target sprite and impact feedback conformance

Raise direct sprite target score and impact/explosion feedback together; current impact/explosion static read is 5.8/10.

estimated; cpu 16 runs; 17.6 min wall; 20.4 min CPU Value score: 0.00

Latest self-critical work-block read

The past focused block substantially improved our honesty and repeatability, but only modestly improved player-facing conformance. Challenge stages are now scored with a strict 1/10 baseline and have risen to 3.8/10; that is a real improvement from the strict 2.5/10 baseline, but still far from human-level Galaga conformance. The biggest remaining failures are movement grammar, alien novelty, stage-to-stage challenge progression, and stable audio runtime promotion.

Weakest human-level gaps: Audio runtime promotion success: 0.0/10; Challenge stage-to-stage progression: 3.0/10; Challenge movement / trajectory conformance: 3.4/10.

Source: reference-artifacts/analyses/conformance-investment-retrospective/latest.json; CONFORMANCE_INVESTMENT_RETROSPECTIVE.md.

Open Cost / Value deep dive Open raw conformance data Open generated release conformance report

Reference ingestion

Ingestion is the repeatable bridge between external artifacts and playable software. It turns reference media into source windows, annotations, scoreable events, visual comparisons, audio cue maps, and measurement debt that future games can reuse rather than rediscovering by taste alone.

Ingestion coverage

Ingestion turns reference media and Aurora runtime captures into repeatable evidence: clips, contact sheets, traces, event logs, labels, scores, confidence, and next missing annotations.

Sources: 16 families; high confidence: 10; scored or promoted: 11.

Galaga target artifact coverage

5.1/10 overall; 4.5/10 challenge-stage readiness. 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.

Critical open sources: 5; missing late challenge windows: 0.

Galaga-family reference audio clips

audio identity / event feedback: 0 clips. clipped, mapped, partially scored; confidence medium-high.

Next: Add finer event labels for explosion, impact, boss damage, immunity/entry, capture, and rescue semantics.

Aurora audio cue comparison and event-gap reports

audio cue scoring: 21 compared cues; semantic 9.78/10; 0 attention rows. semantic-scored; confidence medium-high.

Next: Tune the highest segment-level gap next: challengePerfect onset. Rerun audio comparison and event-gap analysis after the change.

Aurora Audio Conformance Lab v2

audio candidate loop / family promotion decisions: 8/8 target cues swept; 2 keeper candidates tracked; runtime promotions 0; rejected runtime trials 3. family-scored; confidence medium-high.

Next: challengePerfect: Do not promote Challenge Perfect from isolated onset/body candidates. Replace the next audio strategy with full-phrase/segment-boundary work: stabilize the scorer on canonical reference-vs-reference capture, then generate candidates that optimize onset, body, tail, and live capture segmentation together.

Aurora audio cue contracts

audio semantic contract / theme latitude / promotion safety: 8 contracts; readiness 9.1/10; blocked 7; rejected trials n/a. contract-scored; confidence medium-high.

Next: Keep the calibrated layered playerHit runtime cue; next either refine the residual playerHit tail/body gap with the same calibrated scorer or move effort to stagePulse pressure-bed strategy if user impact per compute looks higher.

Open generated game conformance catalog Open Aurora catalog tables Open generated reference media inventory

Game conformance catalog

The project now treats alien/enemy roles, audio cues, stage behavior, and test personas as visible documentation surfaces. The rows below are generated from the live conformance dashboard game profiles; the generated project and application guides carry the fuller tables.

Aurora Galactica

Active conformance investment. Aurora is the shipped first Platinum game and the current subject of long-cycle conformance scoring.

Key gates: Overall quality: 8.8/10; Audio identity: 7.3/10; Challenge-stage set-piece conformance: 4.3/10; Direct target sprite conformance: 6/10.

Next work: Challenge-stage set-piece conformance: movement, graphics, alien novelty (4.3/10); Audio identity, event feedback, and cue alignment (7.3/10); Alien entry and broad challenge-stage novelty (8.2/10).

Live artifacts: src/assets/reference-audio; reference-artifacts/analyses/aurora-audio-event-gap/2026-05-16-08c327dd-dirty-155303/report.json.

Galaxy Guardians

Preview / ingestion. Useful playable-preview evidence now extends beyond the one-level public slice. The score table, wave progression, visible attract/score-table surfaces, frame-motion proxy, object-track proxy, runtime-vs-reference movement comparison, longer-surface stage-band review, persona review runs, stage-rank pressure contract, sprite-reference extraction, component sprite targets, waveform/spectrogram audio comparison, stricter measured cue targets, labeled cue-window previews, reusable audio conformance lab, playtest-weighted conformance review, opening-slice motion targets, combat-feedback frame authority, and platform-frame parity contract now exist as non-production preview contracts. The best-return next work is now deeper stage 3-5 and 6-9 fairness review, a browser-side audio/listening pass, and public-readiness work that extends beyond the current one-level mission_complete slice.

Key gates: Preview reference conformance: 7.6/10; Playtest-weighted conformance: 6.9/10; Public release readiness: 4.2/10; Audio conformance lab: 8.5/10.

Next work: Preview reference conformance (7.6/10); Playable-preview conformance and feel (6.9/10); Audio lab cue identity (8.5/10).

Live artifacts: reference-artifacts/analyses/galaxy-guardians-identity; reference-artifacts/analyses/audio-conformance-lab/galaxy-guardians-preview/audio-conformance-lab-0.1.json.

Confidence rule

An x/10 score is a current-resolution measurement. Low-confidence or proxy rows are deliberately marked so the next work can improve the scorer rather than hide the uncertainty.

Read full generated catalog Open stage summary tables Open platform persona contract

Challenge-stage critical analysis

Challenge stages are being treated as a first-class Galaga conformance problem, not a generic level-arc footnote. The current analysis separates the rule baseline from the richer player experience: challenging stages should not shoot or kill the ship, but they also need memorable alien movement, visible arrival and exit patterns, alien-family variation, and high-bonus learning opportunities.

Challenge-stage critical read

4.3/10 interesting factor and 4.3/10 conformance. 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.

A player should experience challenging stages as safe but tense score exhibitions with memorable entry routes, fresh alien types, readable trajectories, and a learnable perfect-bonus opportunity. Aurora currently preserves the safety rule, but the actual spectacle, motion, and visual novelty are still early.

Challenging Stage 3-4: 4.2/10 interest

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.

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.

Internal marker: 3 Path: first-challenge-peel Best ref: challenge-1-arrival-group-1 (8.5/10)
Movement

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.

Aliens / Graphics

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.

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: 8.6/10. 5/5 wave type signatures and 3/5 path signatures; average within-wave spawn span 0.54s.

Gap / Next

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.

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.

Challenging Stage 7-8: 4.2/10 interest

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.

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

Internal marker: 7 Path: cross-sweep Best ref: challenge-2-arrival-group-1 (8.7/10)
Movement

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.

Aliens / Graphics

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.

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: 8.6/10. 5/5 wave type signatures and 3/5 path signatures; average within-wave spawn span 0.53s.

Gap / Next

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.

Tune challenge 2 toward the denser mixed-novelty-line reference instead of relying on a generic cross-sweep.

Challenging Stage 11-12: 4.4/10 interest

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.

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

Internal marker: 11 Path: hook-arc Best ref: challenge-3-arrival-group-1 (8.5/10)
Movement

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.

Aliens / Graphics

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.

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: 8/10. 5/5 wave type signatures and 3/5 path signatures; average within-wave spawn span 0.36s.

Gap / Next

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.

Promote challenge-3 boss-led reference phases and score dragonfly visual novelty as animation, not only family label.

Challenging Stage 15-16: 4.3/10 interest

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.

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

Internal marker: 15 Path: pink-serpentine Best ref: challenge-4-pink-serpentine-group-1 (8.9/10)
Movement

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.

Aliens / Graphics

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.

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: 7.4/10. 5/5 wave type signatures and 3/5 path signatures; average within-wave spawn span 0.24s.

Gap / Next

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.

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.

Challenging Stage 19-20: 4.4/10 interest

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.

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

Internal marker: 19 Path: pink-green-cascade Best ref: challenge-5-pink-green-cascade-group-2 (9.6/10)
Movement

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.

Aliens / Graphics

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.

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: 7.5/10. 5/5 wave type signatures and 3/5 path signatures; average within-wave spawn span 0.25s.

Gap / Next

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.

Promote the Challenge 5 pink/green cascade window into five group labels and tune lower-field pass timing against those labels.

Open living effort summary Open rendered report Open challenge-stage deep dive Open stage summary Open conformance dashboard

Latest release note

1.4.1 production quality release

Promotes the accepted 1.4.1 beta review as a production quality release with Stage 3 challenge-stage gameplay keepers, beta sign-in repair, audio/theme lane clarity, public-safe foreground audio balance, refreshed security evidence, and stronger release gates while keeping score submission guarded pending server-side validation.

Open release notes Open source markdown

Architecture and code review learning

The project treats review as a learning loop, not just a gate. Architecture review notes, code-review findings, proposed changes, accepted changes, and repeated patterns are captured in a durable ledger so future releases can avoid the same classes of mistakes and decide when a review note deserves a harness, release preflight, platform rule, or documentation rule.

Review cycles 3 Architecture and code-review cycles represented in the learning ledger.
Issue notes 6 Tracked findings, observations, and in-progress review concerns.
Accepted changes 3 Review-driven process or implementation changes accepted into the project.
Blocking findings P0 0 / P1 0 P0/P1 findings block hosted-dev movement. Production dispositions: 3 addressed, 3 dismissed, 0 missing.
P2 - architecture-boundary - architect-key-abstractions

Keep core runtime, shell, service adapter, game-pack, entity, scoring, stage, render, and auth concerns named and separated.

Proposed: Continue moving shared behavior into Platinum-owned contracts and game-specific behavior into pack-owned data or adapters.

Production disposition: addressed - The platform/application architecture documents and boundary harnesses now define the core runtime, shell, service adapter, game-pack, entity, scoring, stage, render, and auth concerns.

Status: accepted Source: ARCHITECT_REVIEW_RESPONSE.md Review decision: Architecture docs and boundary harnesses are the standing source of truth.
P2 - architecture-boundary - architect-magic-constants

Some movement, render spacing, animation timing, and gameplay update values still need clearer structural ownership.

Proposed: Move stable game-design values into pack/spec structures; keep local math values near implementation with intent comments.

Production disposition: dismissed - Remaining movement, render-spacing, animation-timing, and gameplay tuning constants are a next-cycle maintainability concern, not a blocker for the current production candidate when targeted gameplay and boundary checks pass.

Status: in_progress Source: ARCHITECT_REVIEW_RESPONSE.md Review decision: Track this as continuing 1.4.0 platform maturity work.
P2 - harness-fragility - architect-harness-fragility

Recorded-path harnesses can become brittle after motion and pacing changes.

Proposed: Prefer state-based assertions, coarse behavioral envelopes, and repeated-run averages where exact replay geometry is not the contract.

Production disposition: dismissed - Recorded-path fragility is acknowledged and retained as a harness-maturity concern. It does not block the current production candidate if the relevant release harnesses, documentation freshness check, and publish preflight pass.

Status: in_progress Source: ARCHITECT_REVIEW_RESPONSE.md Review decision: Measured conformance work should leave more reusable harness logic behind than one-off replay checks.
P2 - conformance-evidence - architect-enemy-spec

Enemy family definitions should become declarative enough to support new game ingestion and controlled experimentation.

Proposed: Continue extracting enemy family, attack pattern, and stage grammar data into game-owned specs.

Production disposition: dismissed - Declarative enemy-family and attack-pattern specs remain important for 1.4.0 and later conformance work. This is not a production blocker for a reviewed candidate unless the release claims those specs as complete.

Status: in_progress Source: ARCHITECT_REVIEW_RESPONSE.md Review decision: Use ingestion-backed specs before considering heavier persistence infrastructure.
P2 - architecture-boundary - code-review-platform-game-boundary-review

src/js/13-gameplay-adapter-registry.js: Gameplay or pack registry changed; run platform/game boundary harnesses and review pack ownership.

Proposed: Verify the flagged surface with the targeted review questions and recommended checks before lane movement.

Production disposition: addressed - The flagged platform/game boundary surface was reviewed with the targeted boundary harnesses. Aurora remains the active shipped adapter, Galaxy Guardians remains preview/dev-only, and pack registry ownership checks pass.

Status: review_note Source: reference-artifacts/analyses/code-review/latest.json Review decision: Accepted as a visible review note that does not block hosted-dev by itself.
P2 - release-tooling - code-review-release-tooling-review

package.json: Release or npm script surface changed; verify lane authority and publish behavior.

Proposed: Verify the flagged surface with the targeted review questions and recommended checks before lane movement.

Production disposition: addressed - The release-tooling changes are intentional: they add the review packet, review-learning ledger, public review documentation, and a production disposition gate. The review packet is current and publish preflight remains green for the dev lane.

Status: review_note Source: reference-artifacts/analyses/code-review/latest.json Review decision: Accepted as a visible review note that does not block hosted-dev by itself.

release-tooling-churn

Changes to scripts or release tooling often look small but affect lane authority and publish behavior.

Response: Require `npm run publish:check:dev` plus the review packet before hosted-dev movement.

generated-artifact-self-noise

Generated review artifacts can make review packets appear stale if they are treated as source inputs.

Response: Exclude generated review artifact directories from packet freshness while keeping human-facing review docs in scope.

harness-before-subjective-tuning

Motion, audio, pacing, and transition polish can drift when tuned subjectively.

Response: Start with reference evidence, extracted clips, timing windows, or semantic logs, then promote only measured keepers.

review-notes-should-teach

One-off review comments lose value unless they become future prevention mechanisms.

Response: Classify recurring findings and decide whether each should become a harness, release check, documentation rule, or code simplification.

Open review learning ledger Open code review model Open architecture baseline review

Documentation provenance

The main pages are generated from persistent repository artifacts and analysis outputs. This section is itself generated from documentation-provenance.json, so release review can see which artifacts own the visible claims instead of treating page text as the source of truth.

Public Project Page

Narrative public overview of the release lane, conformance program, resource economics, ingestion, and game catalog.

Generated as: public-project-page.html.

Builder: tools/build/build-index.js#buildPublicProjectPage Sections: Release identity and current focus; Conformance readout and release gates; Value versus compute; Reference ingestion Persistent Inputs: release-manifest.json; release-dashboard.json; release-notes.json; reference-artifacts/analyses/release-conformance-dashboard/latest.json
Hosted Project Guide

Long-form hosted guide that renders maintained project sections plus source Markdown documents.

Generated as: project-guide.html.

Builder: tools/build/build-index.js#buildProjectGuide Sections: Guide overview and source document index; Documentation provenance Persistent Inputs: project-guide.json; PROJECT_STATE_AND_CONFORMANCE_PROGRAM.md; CONFORMANCE_INVESTMENT_RETROSPECTIVE.md; CHALLENGE_STAGE_CONFORMANCE_ANALYSIS.md
Hosted White Paper

Hosted narrative surface for the living white paper, its citation ledger, and the supporting white-paper project area.

Generated as: white-paper.html.

Builder: tools/build/build-index.js#buildWhitePaperGuide Sections: White-paper overview and project-area framing; White-paper release memory Persistent Inputs: white-paper.json; WHITE_PAPER.md; white-paper/README.md; white-paper/ILLUSTRATION_PLAN.md
Hosted Release Notes

Hosted readable release-note archive generated from structured release notes plus committed detailed release-note source documents when available.

Generated as: release-notes.html.

Builder: tools/build/build-index.js#buildReleaseNotesPage Sections: Release-note index and latest release framing; Detailed release-note documents Persistent Inputs: release-notes.json; release-dashboard.json; release-manifest.json; RELEASE_NOTE_1.0.0_LAUNCH.md
Aurora Application Guide

Aurora-owned audio, visual, alien, stage, target-artifact, artifact-conformance, and persona catalog.

Generated as: application-guide.html.

Builder: tools/build/build-index.js#buildApplicationGuide Sections: Audio and visual catalog; Artifact conformance status Persistent Inputs: application-guide.json; src/js/13-aurora-game-pack.js; reference-artifacts/analyses/aurora-audio-conformance-lab-v2/latest.json; reference-artifacts/analyses/aurora-audio-event-gap/latest.json
Conformance Dashboard

Primary live score, confidence, cost, ingestion, and investment dashboard.

Generated as: conformance-dashboard.html.

Builder: tools/harness/build-dev-conformance-dashboard-page.js Sections: Dashboard data and metrics Persistent Inputs: reference-artifacts/analyses/release-conformance-dashboard/latest.json

Where to look

Open this lane build Open beta build Open release dashboard Open release notes Open conformance dashboard Open project guide Open application guide Open Platinum guide Open repository Plan source Project state and conformance program Documentation reading paths Guardians resume state Guardians stage assessment Guardians opening baseline Guardians top 10 queue Game conformance catalog Strategic beta review Classic arcade ingestion framework Artifact policy

Development lane project-page summary generated from lane build artifacts. Date label: June 12, 2026.