feat(v1.9.0): Ambition — break the prototype plateau (dual bars, benchmark anchors, technique menu, mega-leap)#74
Merged
Conversation
…, mega-leap, asset ceilings The f1 probe still looked 1980s after v1.7/v1.8: an independent re-grade vs REAL racing games scored 0.09 (F+) while the internal rubric said 0.687 "A". The loop's ceiling is min(standard, medium, leap-scope) — all three were pinned to "prototype". v1.9.0 unpins them: - Dual thresholds bar_leap/bar_coast (rubric_score + loop_scorecard): below bar_leap always leap; coast only above bar_coast (~0.85, anchored to a real product); the forcing zone between keeps leaping. Back-compat: lone `bar` sets both equal. - Benchmark anchors: each dimension carries reference score_0.10..0.90 naming real products; the 5-G critic scores against them, not the artifact's own past (a flat prototype reads ~0.10, not 0.6). prototype_ceiling → ANCHOR WARNING. - Technique menu per dimension (techniques + technique_cdns) injected into the leap so it reaches for EffectComposer/PMREM/particles, not more BoxGeometry. - Mega-leap: human-approved multi-cycle re-platform (bigger budget, atomic rollback) for radical rewrites a bounded leap can't make. - Asset ceilings: ceiling_without_assets → a human_required skill_gap instead of thrashing when code-only work tops out. verify.py 61 → 62. plugin 1.8.1→1.9.0, config schema 1.4→1.5. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The f1 probe still looked like a 1980s game after v1.7/v1.8. An independent re-grade against real racing games scored it 0.09 (F+) — vs the internal rubric's 0.687 "A". A 6-agent diagnosis: the loop's quality ceiling is
min(standard, medium, leap-scope), and all three were pinned to "prototype".Changes
bar_leap/bar_coast— coast only above ~0.85 (anchored to a real product); the forcing zone keeps leaping so the loop can't declare victory at prototype quality. Back-compat preserved.score_0.10..0.90in REAL products; the 5-G critic scores against those, not the artifact's past → a flat prototype reads ~0.10, not 0.6.techniques+technique_cdns(EffectComposer/PMREM/particles…) so they reach for modern techniques, not more BoxGeometry.ceiling_without_assets→human_requiredskill_gap instead of thrashing.verify.py 62 green. plugin 1.8.1→1.9.0, schema 1.4→1.5.
🤖 Generated with Claude Code