feat(submodule): wire Pmoves-pretext — the layout half of the living-docs animation combo - #2227
Merged
Merged
Conversation
…docs animation combo The Remotion+Pretext combo exists to give the living docs animated qualities (CREATOR_PIPELINE.md §2a): Remotion is the motion runtime in the A2UI Renderer, Pretext is the deterministic text-layout engine that keeps captions/overlays/living-doc copy stable in exported motion assets. The tour claim audit (2026-07-25) found CREATOR_PIPELINE.md's 'PMOVES tracks the Pmoves-pretext fork' was not structurally true — the fork existed on GitHub but was never wired in. Mechanical wiring per fleet convention (OpenRoom precedent): - PMOVES.AI-Edition-Hardened minted on the fork @ upstream main bb224e08 (0 ahead / 0 behind chenglou/pretext) - submodule added tracking hardened, ignore=all - skill registry entry (tier 4, ui) - submodules.md catalog entry with the real integration points and the render proof: the living-doc animation lane was exercised end-to-end today — ProvenanceLivingDoc composition rendered via render:provenance (411-frame 30fps 1080p MP4 + still) with pretext layout in the loop. Runtime still consumes npm @chenglou/pretext@0.0.6; the submodule is the alignment/patch surface until fork releases replace the dep. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Contributor
|
Caution Review failedThe pull request is closed. ℹ️ Recent review info⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: CHILL Plan: Pro Plus Run ID: 📒 Files selected for processing (4)
📝 WalkthroughWalkthroughAdds the ChangesPmoves-pretext integration
Estimated code review effort: 1 (Trivial) | ~5 minutes Possibly related PRs
Suggested labels: ✨ Finishing Touches🧪 Generate unit tests (beta)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 3106062848
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| url = https://github.com/POWERFULMOVES/PMOVES-OpenRoom.git | ||
| branch = PMOVES.AI-Edition-Hardened | ||
| ignore = all | ||
| [submodule "Pmoves-pretext"] |
There was a problem hiding this comment.
When adding this submodule, the mandated pmoves/docs/AGENTS/SUBMODULE_CODEX_HOMES/README.md workflow requires a matching Pmoves-pretext.md overlay in the same change, but no such file exists. Consequently, codex_submodule_audit.py reports this module without Codex artifacts and provides operators no module-specific routing or validation guidance; add the overlay and regenerate the audit.
AGENTS.md reference: AGENTS.md:L99-L102
Useful? React with 👍 / 👎.
POWERFULMOVES
pushed a commit
that referenced
this pull request
Jul 27, 2026
Two real, different 'pretext's now ship in the fleet and the index must keep them apart: (1) @chenglou/pretext — the deterministic text-LAYOUT engine in the a2ui-renderer Remotion runtime (fork now wired as the Pmoves-pretext submodule, #2227 — row updated from 'not wired'); and (2) PreTeXt (pretextbook.org) — the structured math-AUTHORING language the persona living-doc room lane used to build 'CHIT & the MOF: A Structural Isomorphism' (pmoves/rooms/persona/pretext, buildable with pretext-cli 2.45.0; verified valid XML, 3 sections, 21 math elements). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
This was referenced Jul 27, 2026
POWERFULMOVES
pushed a commit
that referenced
this pull request
Aug 3, 2026
Two real, different 'pretext's now ship in the fleet and the index must keep them apart: (1) @chenglou/pretext — the deterministic text-LAYOUT engine in the a2ui-renderer Remotion runtime (fork now wired as the Pmoves-pretext submodule, #2227 — row updated from 'not wired'); and (2) PreTeXt (pretextbook.org) — the structured math-AUTHORING language the persona living-doc room lane used to build 'CHIT & the MOF: A Structural Isomorphism' (pmoves/rooms/persona/pretext, buildable with pretext-cli 2.45.0; verified valid XML, 3 sections, 21 math elements). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
POWERFULMOVES
added a commit
that referenced
this pull request
Aug 4, 2026
…f gate) (#2076) * feat(dl-1b): CHIT tour re-skin to PMOVES armor tokens + port onto current site Ports the held CHIT Visual Tour (12-commit worktree branch, built pre-DL-1) onto current main and re-skins its AI-default chrome to the armor tokens — the original hold reason was exactly this: a 4th divergent look. Palette: one uniform role mapping applied everywhere (CSS vars, JS string + 0x hex forms, favicon %23 forms): old primary teal -> pm-accent #7C3AED, old primary-soft -> pm-accent-soft #A78BFA, old amber accent -> pm-accent-2 Reserved DARKXSIDE signature #E11D48 added as --signature, applied to ONE element (section markers, large text). Tier-encoding colors (--t-*) and warn/error/success semantics untouched. Zero old-palette hexes remain outside vendor/. Shared-file touches re-applied against current main: nav link + demos card (index.html), sitemap entry (lastmod 2026-07-11), narrow /chit-tour/* CSP block (style-src-attr only; scripts stay strict 'self'). Gate: DARKXSIDE visual sign-off before merge. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * feat(chit-tour): verified-actuals evidence + regulatory anchor + projection honesty Operator review verdict on the tour: no aspirational content presented as fact — verified actuals with evidence, provenance proof, and current regulatory anchoring. - Tokenomics section now leads with MEASURED results: real export run (2026-07-25T14:02Z, baseline, 52 weeks, 156 transactions, dry-run) with real week-1 ledger records, the NATS event on the registered tokenism.export.result.v1 subject, and a provenance chain — canonical event payload + sha256 commitment, honestly labeled commitment-not- signature (unsigned demo mode), anchored to PR #2169 and the ToKenism-Multi export sidecar (PR #52). - Regulatory anchor card (status as of 2026-07-25, engineering posture not legal advice): GENIUS Act rulemaking state and the two design rails it locks (no fixed-USD redemption promise, no yield-on- holdings — distributions stay contribution-keyed, which the measured run demonstrates); CLARITY Act Senate status with both classification branches kept priced in; queued simulation updates (attribution-vs- yield tripwire, AML overhead line item, regulatory-timeline scenario variable, reserve model parked). - Every projection visualization now carries a PROJECTION MODEL chip (--pm-signature-soft, AA-safe) and the section lede separates measured results from planning instruments explicitly. - LIVING_DOCS_INDEX: Creator pipeline row now links the actual living doc (CREATOR_PIPELINE.md — Remotion runtime); new rows for the Pretext text-layout fork (Pmoves-pretext, consumed by a2ui-renderer/src/remotion), the CHIT tour itself (with its verified-actuals evidence policy), and the skills registry. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * fix(chit-tour): audit fixes — canon sourcing, archetype count, index accuracy Applies the independent claim audit: - CHIT two-meaning split: keep it (it is brand-owner canon, 2026-06-09 reconciliation of 7 historically conflicting expansions) but source it HONESTLY — new lead coverage note states the PMOVESCHIT files still carry the older single expansion pending the fleet-wide doc sweep, so the split is the canon speaking, not the cited file. - Projection duplication disclosure corrected: SIX archetypes share the byte-identical curve (was undercounted as four), verified row-by-row. - CGP version-currency note: spec v1.0 production-ready while recorded example payloads still show v0.1/v0.2 — spec ahead of examples. - LIVING_DOCS_INDEX accuracy: PMOVES-Creator is the ComfyUI image/render farm (zero Remotion — previous row misattributed); the Remotion runtime is the A2UI Renderer (8107). Pretext row now states the real wiring: npm @chenglou/pretext@0.0.6, fork exists on GitHub but is NOT a submodule. Skills row disambiguates the two registries (submodule map vs PmovesSKillZ operator-workflow skills). Tour row cross-links the sibling code-first VISUAL_TOUR.md so the two stop drifting silently. Audit false-alarm for the record: pmoves/config/rooms/ tokenism.room.exchange.json exists on main (sampling miss) — the W2 evidence citations stand. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * feat(chit-tour): Poincaré section computes real hyperbolic geometry — no more decorative dots Operator verdict on the old section: dots with no relevant info, nothing showing what is flattened from hyperbolic space — 'the math needs to show teeth, not fake diagrams.' The old code seeded jitter at fixed radii with straight-line links; it also drew two 'Legendary' center dots that do not exist in the registry. Replaced with computation, all client-side on load: - REAL data: AGENT_TREE extracted verbatim from pmoves/config/agent_registry.yaml (taxonomy v1.5.0, 96 agents; registry truth: 43 standard / 28 specialized / 24 utility / 1 ci, ZERO legendary — the page says so). - Sarkar's construction (2011), scaled variant tau=[1.1,1.7,2.3]: Mobius-translate each node to the origin, fan children at hyperbolic distance tau facing away from the parent, translate back. - Every edge is a true geodesic: arc of the circle through both points orthogonal to the unit circle; diameters handled as lines. - Equal-hyperbolic-step ruler rings at r=tanh(d/2) — the crowding toward the rim is the flattening cost, stated as such. - Interactive metric: click two nodes → the arcosh distance formula evaluated with their computed coordinates, plus euclidean distance, compression factor, tree-path length, and realisation ratio. - Measured distortion of THIS embedding: mean d_H/d_tree = 0.674, worst-case contraction x21.2 over 393 sampled leaf pairs — with the honest tradeoff explanation (Sarkar's theorem: ratio → 1 as tau grows, at the cost of rim crowding). - 'What is being flattened': hyperboloid sheet x^2+y^2-z^2=-1 with class/type nodes lifted via the inverse stereographic map and projection rays from S=(0,0,-1) to their disk shadows; agent leaves truncated from the 3D view with the reason stated. - Labels escaped before innerHTML (A2UI hardening rule); WebGL loop throttled ~20fps and visibility-gated. Headless verification: embed 0.67ms, 120 nodes, 0 invalid coordinates, 0 NaN geodesic paths; browser DOM probe confirms 256 rendered SVG elements and live metric/distortion panels. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * feat(chit-tour): four pillar laboratories — every pillar's math computed live Completes the operator's 'math needs to show teeth' directive across all five pillars (Poincaré/Pillar 2 landed in the previous commit). Each lab runs the pillar's actual mathematics in-browser with the exact parameters the reference implementations use, and states honestly where shipped code and doc prose diverge. Pillar 1 — Dirichlet (dirichlet-weights.ts): live Dir(α) sampling (Marsaglia–Tsang gammas) scattered on the 2-simplex; the EXACT update rule α_i += amount·concentrationK with the real defaults (smoothingAlpha 0.1, concentrationK 1.0); E[w]=α/Σα marked; the non-zero-share guarantee visible in every draw. Pillar 3 — Merkle (shape-attribution.ts): four REAL leaves — the three week-1 ledger transactions and the result event from the measured 2026-07-25 export run — hashed with WebCrypto SHA-256 in the ts leaf shape {address, action, amount, week, category}; click-to-prove inclusion {path, pathIndices} verified against the committed root; a tamper button edits one amount and the proof visibly dies. Pillar 4 — Zeta (zeta-filter.ts): the 20 zero constants verbatim, weights w_n = decayⁿ/ln(γ_n) normalized, applied as the same circular convolution filterSpectrum() implements; REAL input = the 4-bin CGP spectrum from the worked example in 01_WHAT_IS_CHIT.md; a second 32-bin input is labeled synthetic; N and decay are sliders with the ts defaults (10, 0.9). Pillar 5 — Swarm (swarm-attribution.ts): honesty note front and center — the ts header states it does NOT perform mutation/selection/ crossover; it is the fitness scorer + population tracker. Panel A runs the real scorer with the exact gini_reduction weights (0.6/0.2/0.1/ 0.05/0.025/0.025; targets Gini 0.3, poverty 0.1). Panel B is a clearly-labeled demo of the doc-described evolutionary loop driving that same scorer, with real Gini computed per generation. Also: hero/KPI agent counts corrected to the registry truth (96, not the taxonomy doc's stale 76) with a coverage note recording the doc lag. Headless verification: Dirichlet draws sum to 1 with all components positive; zeta weights normalize to 1 and reproduce filterSpectrum on the doc spectrum; fitness monotonic vs Gini; Gini 0 on equal / 0.75 on fully-skewed allocations. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * feat(chit-tour): digest strips for every pillar lab + the tokenizable-by-design convergence Two operator directives: 1. Digestibility — every pillar laboratory (including the Poincaré section) now opens with a three-line plain-language strip: 'Plain words' (what this machine is, no jargon), 'Try this' (one concrete interaction and what you'll see), and 'Why it matters for attribution' (the stake). The math keeps its teeth below; the strip is the on-ramp. 2. Shape attribution is tokenizable BY DESIGN — new convergence block closing the pillars section, mapping the five labs onto the single pipeline stage by stage: domain content → CGP shape (P4 filter + P2 hierarchy) → Dirichlet shares (P1 — the split IS the cap table) → Merkle-committed records (P3 — the shipped contract already carries token_received / reward_claimed action types and verifies claims against the root, shape-attribution.ts:404) → swarm-scored allocation pressure (P5) → the measured ToKenism ledger path (§10, dry-run gated). The interop consequence stated plainly: external domains connect by emitting a CGP for THEIR domain — their constellations get shares, their shares get proofs, their claims verify against the same roots. Evidence it already happens: the live registry carries 18 external contributors (hermes-agent, claude-opus, kilocode, codex, …) beside the 96 internal agents. Regulatory rails from §10 hold by construction: contribution-keyed shares, proofs not redemption promises. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * feat(chit-tour): planned regulatory sims EXECUTED — results replace plans The regulatory anchor's queued simulations have been run (2026-07-26); the card now reports results, not intentions: 1. Attribution-vs-yield tripwire — REAL, full corpus: all 156 transactions of a 52-week baseline run checked (52 pool transfers + 104 external-spending withdrawals). Zero yield/interest/APY-shaped flows: every distribution is contribution-keyed, which is the GENIUS-era design rail demonstrated on actual output. Self-test proves the checker flags a synthetic yield-on-holdings transaction. (First attempt vacuously passed over 0 transactions — extraction missed the sim shape; caught and fixed before publishing. The transformer path is transform_simulation_to_transactions over the full history.) 2. AML/BSA overhead on breakeven — model, labeled ,500/mo program assumption. Honest headline: NO archetype survives standalone (best implied monthly net ~,064 < program cost). Pooled across container replicas: 3 → 8.9mo, 5 → 6.5mo, 10 → 5.5mo breakeven. Under GENIUS-era compliance the Docker-style replication argument is the only viable micro-issuer structure, not an optimization. 3. Regulatory-timeline scenario variable — model, labeled branch factors: GENIUS early/Jan-2027 x CLARITY pass/stall; both-delayed prices a ~22% ambiguity haircut on the top archetype (EV 1455.5 → 1135.3). Reserve-model stays parked (licensed-issuer conversation trigger). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * docs(index): disambiguate the two pretexts + submodule status current Two real, different 'pretext's now ship in the fleet and the index must keep them apart: (1) @chenglou/pretext — the deterministic text-LAYOUT engine in the a2ui-renderer Remotion runtime (fork now wired as the Pmoves-pretext submodule, #2227 — row updated from 'not wired'); and (2) PreTeXt (pretextbook.org) — the structured math-AUTHORING language the persona living-doc room lane used to build 'CHIT & the MOF: A Structural Isomorphism' (pmoves/rooms/persona/pretext, buildable with pretext-cli 2.45.0; verified valid XML, 3 sections, 21 math elements). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> * feat(chit-tour): accuracy pass + CHIT playground demo Tour §10 (Tokenomics): rewrite from the real ToKenism Multi codebase — 4 Python scenarios + 5 TS projection models surfaced, TradFi vs Cooperative weekly ledger with hand-verified math, 3-channel Gini reduction explanation, link to live Tokenism UI (:3000). Tour §10 (Shape attribution): compounding (Dirichlet α additive), math-provable (Merkle inclusion proofs), tangible (real $1,880.71 records), full API surface (7 action types, verifyProof, getHyperbolicEncoding, exportCGP), honesty note on convergence. Tour §10 (Real-world context): Ioniq 5 reframed as PMOVES mobile fleet infrastructure (GCEW/OJET EV grants). MWBE/SDVOB certifications. Fordham Hill pilot demographics. More Perfect Union video link. Projection chips relabeled PILOT ECONOMICS. Tour §12 (Coverage): bare file paths replaced with per-pillar explanations of what each source proves. New cross-link panel: 3D geometry ↔ Tokenism economics ↔ real-world fleet context. CHIT Playground (website/chit-demo/): standalone interactive demo with real WebCrypto HMAC-SHA256 signing, 4 real repo scenarios (consciousness theories, urban farming, cooperative economics, agent taxonomy), spectrum visualization, signing flow animation, Shape ID computation, and tamper-detection demonstration. --------- Co-authored-by: Shaela Bello <slbello@uncg.edu> Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
POWERFULMOVES
added a commit
that referenced
this pull request
Aug 7, 2026
…, no lanes closed The register carries 121 CLAIM lines against 115 RELEASE lines. Seven CLAIMs dated 2026-07-30 onward have no matching RELEASE. This records, per lane, whether the work it cites actually merged — so the owning agent can close from evidence instead of memory. Deliberately writes ZERO RELEASE lines. Verified: 30 additions, 0 deletions, no "RELEASE `" line added. These are Mavis's and CRUSH's lanes; under Village Rule they sign off their own work. Producing a sign-off on another agent's behalf would make the coordination surface less trustworthy, not more — which is the exact problem this sweep exists to address. Verified shipped (RELEASE appears to be all that is missing): line 1466 CRUSH B850 Knuckles convergence #2288 MERGED line 1552 Lane 2228 a2ui-renderer service-source #2227 MERGED line 1587 Lane 2228 yt-cookies Auth refactor #2327/#2333/#2346 MERGED line 1624 PR #2417 dockerfile-paths ratchet #2415/#2416/#2417 MERGED line 1594 Issue #2217 item 4 Archon auth entry says "closed", no RELEASE Needs an owner answer — no PR cited, nothing merged found: line 1500 Lane 3 supabase-stack-default-up ~7 days open line 1511 Lane 4 test_all_tts_engines -> pterm ~6 days open Probably shipped, owner to confirm: line 1559 Lane 5 NATS broker / void-publish fix no PR cited, but the ARCHON stream landed in #2397 and was verified on 5090, consistent with completion. Not asserted. Method is recorded in the entry so it is reproducible rather than trusted: counts read from origin/main, PR numbers extracted per CLAIM line, merge state via `gh pr view --json state,mergedAt`. Automated CLAIM-to-RELEASE pairing is unreliable here because Mavis publishes every lane under one agent id, so a later RELEASE on any lane masks earlier open ones — hence a verification list, not a reconciliation. Four of the seven are bookkeeping. Lines 1500 and 1511 are the two that matter for stabilization: nothing merged, and nothing records what happened. Out of scope by design: submodule sync + audit (z890-claude's active lane). Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
POWERFULMOVES
added a commit
that referenced
this pull request
Aug 7, 2026
…, no lanes closed (#2465) The register carries 121 CLAIM lines against 115 RELEASE lines. Seven CLAIMs dated 2026-07-30 onward have no matching RELEASE. This records, per lane, whether the work it cites actually merged — so the owning agent can close from evidence instead of memory. Deliberately writes ZERO RELEASE lines. Verified: 30 additions, 0 deletions, no "RELEASE `" line added. These are Mavis's and CRUSH's lanes; under Village Rule they sign off their own work. Producing a sign-off on another agent's behalf would make the coordination surface less trustworthy, not more — which is the exact problem this sweep exists to address. Verified shipped (RELEASE appears to be all that is missing): line 1466 CRUSH B850 Knuckles convergence #2288 MERGED line 1552 Lane 2228 a2ui-renderer service-source #2227 MERGED line 1587 Lane 2228 yt-cookies Auth refactor #2327/#2333/#2346 MERGED line 1624 PR #2417 dockerfile-paths ratchet #2415/#2416/#2417 MERGED line 1594 Issue #2217 item 4 Archon auth entry says "closed", no RELEASE Needs an owner answer — no PR cited, nothing merged found: line 1500 Lane 3 supabase-stack-default-up ~7 days open line 1511 Lane 4 test_all_tts_engines -> pterm ~6 days open Probably shipped, owner to confirm: line 1559 Lane 5 NATS broker / void-publish fix no PR cited, but the ARCHON stream landed in #2397 and was verified on 5090, consistent with completion. Not asserted. Method is recorded in the entry so it is reproducible rather than trusted: counts read from origin/main, PR numbers extracted per CLAIM line, merge state via `gh pr view --json state,mergedAt`. Automated CLAIM-to-RELEASE pairing is unreliable here because Mavis publishes every lane under one agent id, so a later RELEASE on any lane masks earlier open ones — hence a verification list, not a reconciliation. Four of the seven are bookkeeping. Lines 1500 and 1511 are the two that matter for stabilization: nothing merged, and nothing records what happened. Out of scope by design: submodule sync + audit (z890-claude's active lane). Co-authored-by: Claude Opus 5 (1M context) <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
2 participants
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.
Operator direction: the Remotion+Pretext combo's whole purpose is giving the docs animated living qualities — and the audit found the Pretext half was claimed (
CREATOR_PIPELINE.md'PMOVES tracks the fork') but never structurally wired.Mechanical wiring (OpenRoom precedent): hardened branch minted on
Pmoves-pretext@ upstream mainbb224e08; submodule tracking it (ignore=all); skill-registry entry;submodules.mdcatalog entry documenting the real integration points.The lane itself is proven live (2026-07-25): deps installed in
pmoves/services/a2ui-rendererand the living-doc animation rendered end-to-end —ProvenanceLivingDoccomposition viarender:provenance, 411 frames @ 30fps, 1920×1080 MP4 + still, pretext layout in the loop. That's the 'docs with animated living qualities' demonstration, produced by the actual pipeline.Known follow-ups (not this PR): the renderer has NO compose service —
CREATOR_PIPELINE.mddocuments port 8107 but nothing deploys it (Dockerfile exists; needs a compose stanza + make target soPOST /render/provenanceis a fleet surface, not an npm script); runtime still consumes npm@chenglou/pretext@0.0.6— the submodule is the patch/alignment surface until fork releases replace the dep.🤖 Generated with Claude Code
Summary by CodeRabbit