@chrono-meta/fh-gate 1.4.73 → 1.4.75

package/knowledge/shared/patterns/multi-persona-review.md

@@ -0,0 +1,88 @@
+---
+name: multi-persona-review (parallax)
+description: Generalized architecture for multi-persona parallel artifact review — parallel isolated personas + shared output protocol + neutral synthesizer. The domain-agnostic, IP-stripped pattern named "parallax" for FH; embodied as sim-conductor Step 1.5. Use when designing or explaining structured multi-viewpoint review (code, spec, design, plan).
+type: reference
+tags: [pattern, multi-persona, review, parallax, sim-conductor, verify-axis]
+---
+
+# Multi-Persona Parallel Review — the `parallax` pattern
+
+> **parallax**: depth — the hidden, the omitted — is only perceivable from multiple viewpoints. One
+> standpoint is flat; the shift *between* standpoints reveals what a single seat cannot. The name is the
+> thesis: structured multi-standpoint coverage finds categorically more than any one viewpoint.
+
+**Provenance**: generalized and domain-stripped from a field marketplace implementation (`deep-insight`,
+a company-internal term — not adopted here). The domain-agnostic *architecture* is portable; the
+domain-specific personas were field extensions and are excluded. In FH this pattern is a **mode of
+`sim-conductor`** (Step 1.5), not a separate skill — see `asset-placement-gate` verdict (2026-06-06):
+it overlaps sim-conductor ~80–90%, so the distinctive patterns were absorbed, not forked.
+
+## The architecture (6 + 1 patterns)
+
+1. **Parallel isolated persona spawn** — N personas run as parallel agents (`Task(subagent_type=…)` /
+   parallel Agents), each in its own context, no inter-agent communication. Isolation is mechanical: it
+   lets each persona reach an independent verdict, which is what makes cross-persona agreement meaningful.
+2. **Shared output protocol (constitution)** — every persona, whatever its lens, emits the same shape:
+
+   ```
+   ### Strengths        (0–3, from this persona's viewpoint)
+   ### Concerns
+     Critical   — compile/runtime failure · clear logic error · data corruption · security leak
+     Important  — significant user/service impact in a plausible scenario
+     Suggestion — optional improvement
+     (each: [file:line or quoted span] one-line summary — rationale)
+   ### Open questions   (0–3 needed for a decision)
+   ### Absence check    (outside-vantage personas: what does the artifact FAIL to specify that this
+                         standpoint needs — discoverability, undocumented contract, unstated assumption?)
+   ```
+3. **Neutral synthesizer** — a NON-persona aggregator that injects no opinion: never a conclusion no
+   persona stated · preserve attribution (who said what) · priority labels verbatim · no forced consensus
+   or forced conflict. It reports Common opinions (2+ agree) and Conflicts (A vs B, each with rationale).
+4. **Persona taxonomy (group aliases)** — `dev` (fe/be/…), `biz` (pm/designer/…), `review`
+   (qa/compliance/devil-advocate), `user` (newcomer/power-user), `all`. Sourced, not hardcoded.
+5. **FP judgment discipline** — only escalate when confident. Never escalate: pre-existing issues ·
+   style without a quotable rule · linter-catchable · speculative · subjective (→ Suggestion). *If not
+   confident, do not mark it* — false positives erode reviewer trust.
+6. **Devil's advocate as a structural member** — not a special case; a standard member of `review`. Every
+   objection carries a rationale (no contrarianism for its own sake); the faster others converge, the
+   harder it hunts the weak point.
+7. **External-harness persona sourcing** (FH addition) — a lens may be sourced from an installed *sibling
+   harness*, not only the built-in palette: e.g. gstack `/review` (staff-engineer), `/cso`
+   (security-officer), `/qa` (QA-lead). The orchestrator runs FH; the sibling supplies the specialist
+   lens. Isomorphic to `steel-quench` Step 0.4 (Specialized Reviewer Discovery) + Wave 5 (external CLIs).
+
+## The two-layer severity (why it's required, not redundant)
+
+Per-persona `Critical/Important/Suggestion` (each persona's own judgment in isolation) → synthesizer
+triages to `M/S/R` (cross-persona). An isolated persona **cannot** assign `S = found by 3+ personas` —
+that depends on agreement it never sees. So the per-persona → synthesized split is *required by isolation*,
+not duplicate vocabulary.
+
+## What this pattern's value actually IS (honest, from the R9 experiment arc)
+
+Established by controlled experiment (companion-store `RESULT9`, codex, capability-controlled, 2026-06-06):
+
+- **It is NOT de-biasing.** Ownership had ≈ zero effect (a model self-reviewing blunt = an outsider
+  blunt). It is NOT "seeing what the author structurally can't": when explicitly asked, an author finds
+  their own omissions about as well as an outsider (clean replication refuted the "non-self-administrable"
+  claim).
+- **What is real**: the gradient **`blunt < single-rubric < multi-persona`** — structured multi-standpoint
+  coverage finds categorically more (a single-flaw prompt missed SSRF; a single rubric missed the
+  consumer-contract axis a newcomer surfaces). Different standpoints make different issues *visible*.
+- **So the value is**: **pure rubric/standpoint supply + routine enforcement** — the pattern assembles and
+  *routinely runs* the structured multi-standpoint rubric (including the omission/absence check) that a
+  builder would not assemble or run on themselves. **Fully copyable utility, zero cognitive-moat.**
+  Isolation's role is mechanical (independent verdicts → honest cross-persona triage + decorrelation), not
+  cognitive.
+
+This framing is falsifiable and matches FH's utility-not-moat positioning: the personas are a copyable
+marketplace; the marginal value is making the structured sweep routine, not exclusivity.
+
+## Where it lives in FH
+
+- **Operational home**: `sim-conductor` Step 1.5 (Persona Output Protocol + Neutral Synthesizer). Area D
+  (artifact validation) is the focused multi-persona review entry point.
+- **Sourcing kinship**: `steel-quench` Step 0.4 + Wave 5 (external reviewer/CLI discovery).
+- **6-axis**: a validated pattern for **Axis 5 (Verify)** — structured multi-viewpoint review of a
+  near-complete artifact, complementary to steel-quench (adversarial hardening) and phantom-quench
+  (source grounding).

package/knowledge/shared/plugin-catalog/recommended_plugins.md

@@ -0,0 +1,117 @@
+---
+name: forge-harness Recommended Plugins Catalog
+description: Persistent record of plugin-recommender skill execution results. The primary asset for plugin-recommender skill Step 2 [1st priority] Curated List. Updated cumulatively each time the skill runs.
+date: 2026-05-08
+version: 0.1
+---
+
+# forge-harness Recommended Plugins Catalog
+
+> Primary search asset for the plugin-recommender skill. Updated cumulatively each run.
+
+## Run History
+
+| Run | Date | Context | Result |
+|:--:|---|---|---|
+| 1 | 2026-05-07/08 | `.claudeignore` + model switching / project automation path | Organization GHE 0 items / Claude Code built-in sufficient / external OSS simulation phase |
+| 2 | 2026-05-08 | UX Writing QA / NLP — self-simulation | Top asset utilization + md-wiki activation / external NLP plugin simulation phase |
+| 3 | 2026-05-08 | Virtual: full-stack (React + Spring Boot) — self-simulation | `.claudeignore` big effect (node_modules·target·build) + model switching / external LSP·ESLint simulation phase |
+| 4 | 2026-05-08 | **Organization GHE Tier 1 cross-recommendation simulation** — 5 repos | ★★★ 2 items + ★★ 3 items found / harness return value 4 paths |
+| 5 | 2026-05-08 | **Organization GHE Tier 2 sister asset cross-recommendation simulation** — 6 repos | ★★★ 1 item + ★★ 4 items + ★ 2 items / Tier 1+2 cumulative n=11 |
+
+## Category 0.5 — claude-plugins-official (Tier 1 official — check BEFORE building or searching elsewhere)
+
+> **No-reinvention rule (meta-harness 철칙)**: if an official plugin covers the capability, use it —
+> do not rebuild it as an FH skill or hunt third-party. FH builds only what adds *governance* on top.
+> Source: [`anthropics/claude-plugins-official/plugins`](https://github.com/anthropics/claude-plugins-official/tree/main/plugins)
+> — inventory snapshot 2026-06-10 (36 plugins), **name-based grouping; verify each plugin's actual
+> behavior via its README before install — do not trust the group label alone. Re-enumerate the
+> directory when recommending; this list goes stale.**
+
+| Group | Plugins | FH / mapped-project relevance |
+|---|---|---|
+| **Language servers (12)** | clangd · csharp · gopls · jdtls · kotlin · lua · php · pyright · ruby · rust-analyzer · swift · typescript (`*-lsp`) | Full-Harness Mode step 4: match to the mapped project's language — instant capability boost, zero FH build |
+| **Dev workflow** | feature-dev · commit-commands · pr-review-toolkit · code-review · code-simplifier · code-modernization · frontend-design | Field-project execution layer. Role split: code work → these; FH-asset coherence → FH skills |
+| **Authoring/extension** | skill-creator · plugin-dev · hookify · mcp-server-dev · agent-sdk-dev | Drafting tools. **FH 6-item gate still applies to anything they produce** |
+| **Setup/session** | claude-code-setup · claude-md-management · mcp-tunnels · session-report · explanatory/learning-output-style | ⚠️ `claude-md-management` overlaps FH territory — role split: file mechanics → plugin; methodology/governance → FH. Evaluate before adopting |
+| **Other** | security-guidance · ralph-loop · math-olympiad · playground · example-plugin · cwc-makers | `security-guidance`: candidate complement to Pre-Publish gate — evaluate |
+
+## Category 1 — Claude Code Built-in (Standard usage / no install)
+
+| Item | Mechanism | Usage path |
+|---|---|---|
+| **`.claudeignore`** | `.gitignore` syntax / per-session token cost 20-40% ↓ | `cp templates/.claudeignore <project>/.claudeignore` |
+| **`/model` + model switching** | Switch models mid-session | `/model opus` / `/model sonnet` / `/model haiku` |
+
+## Category 2 — Forge-Harness Meta Skills (already installed / use fully)
+
+- **fh-meta 5 skills**: `harvest-loop` · `verify-bidirectional` · `frontier-digest` · `cross-ecosystem-synergy-detection` · `plugin-recommender`
+- **`hub-persona-auditor`** + **`fact-checker`** agents
+- **Active onboarding protocol** (forge-harness `CLAUDE.md`) — greeting/start trigger → 5-skill cascade
+
+## Category 3 — Organization Internal Tools / MCP (install as needed)
+
+> Replace this section with your organization's specific MCP tools and internal plugins.
+
+| Item | Source | Usage path |
+|---|---|---|
+| **`<your-mcp-plugin>`** | `<your-org>/<your-plugin-repo>` | Configure in `.mcp.json` / refer to your organization's MCP setup guide |
+
+### Examples of what to put here:
+- Calendar / task management MCP (e.g., Google Calendar, Jira)
+- Internal wiki / docs MCP
+- Slack / messaging MCP
+- Organization-specific API adapters
+
+## Category 4 — External OSS (evaluate then decide / simulation phase)
+
+| Item | Evaluation | Priority |
+|---|---|---|
+| TestSprite MCP | May be covered by internal tools | ⚠ Follow-up |
+| TestCollab MCP | May be covered by internal tools | ⚠ Follow-up |
+| claude-code-skills (levnikolaevich) | Possible overlap with fh-meta 5 skills | ⚠ Follow-up |
+| ClaudeMarketplaces.com (4,200+ skills) | Additional search area — not covered in 1st run | ⚠ Follow-up |
+| awesome-claude-skills (ComposioHQ) | Curated list — not covered in 1st run | ⚠ Follow-up |
+
+## Domain-Specific Simulation Matrix (n=3 accumulated)
+
+| Domain | ★★★ Immediate | ★★ Gradual | ★ Follow-up evaluation |
+|---|---|---|---|
+| **QA automation** (test prep automation) | `.claudeignore` · active onboarding · `verify-bidirectional` | `harvest-loop` | domain-specific test plugins |
+| **Technical writing / docs QA** | `.claudeignore` · active onboarding · `<your-wiki-mcp>` · `verify-bidirectional` | `<your-messaging-mcp>` | NLP plugin |
+| **Virtual full-stack** (React + Spring Boot) | `.claudeignore` (node_modules·target big effect) · active onboarding · `/model opus` for reasoning | `verify-bidirectional` | ESLint MCP · TS LSP |
+
+### Common 3 items (all domains ★★★)
+**`.claudeignore`** + **active onboarding protocol** + **harness meta skills**
+
+→ These 3 are automatically applied in active onboarding Step 4 setup. Core of the user's fast path (direct implementation of harness mission — *"easy + no setup burden"*).
+
+### Domain-differentiated tools
+
+| Domain area | Differentiated tool |
+|---|---|
+| QA domain | `verify-bidirectional` · `harvest-loop` |
+| Frontend/full-stack | `/model opus` (coding/reasoning split) |
+| Wiki/documentation | `<your-wiki-mcp>` (publish · update · sync) |
+| Collaboration notifications | `<your-messaging-mcp>` |
+| All domains | harness meta skills (cross-validation) |
+
+### n=3 simulation learnings (maturity improvement path)
+
+1. **Common 3 auto-applied** — active onboarding protocol naturally manifests (no user burden)
+2. **Domain-differentiated tools matched after user answer** (plugin-recommender 5-step)
+3. **External OSS decided after evaluation** (simplification guard — 80%+ covered by harness meta skills)
+4. **plugin-recommender skill v0.2 patch decision after 3+ runs** (currently at 3 runs)
+
+## Update obligation (cumulative each run)
+
+- Add 1 row to run history table
+- Update per-category matrix (new recommendations / evaluation changes)
+- Persist domain-specific results
+- Immediately reflect when new insights are caught
+
+## Next trigger
+
+- On new work/project entry, read this file first → match against Category 1·2·3 first
+- On external plugin search cycle, update Category 4 (ClaudeMarketplaces.com / awesome-claude-skills direct fetch)
+- plugin-recommender skill v0.2 evolution: decide after 5-run accumulation

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@chrono-meta/fh-gate",
-  "version": "1.4.73",
+  "version": "1.4.75",
   "description": "FH runtime adapters — run FH governance, skills, and agents via Claude or Codex with machine-parseable gates.",
   "license": "MIT",
   "keywords": [
@@ -63,8 +63,29 @@
     "scripts/fh-goal.sh",
     "scripts/count_check.sh",
     "scripts/selfcheck.sh",
+    "scripts/package_coverage_check.sh",
     "scripts/test_fh_gate_regressions.sh",
     "templates/local_fh_context.md",
+    "docs/ETHOS.md",
+    "docs/WHY.md",
+    "docs/OUTPUT_EVIDENCE.md",
+    "knowledge/shared/GLOSSARY.md",
+    "knowledge/shared/patterns",
+    "knowledge/shared/plugin-catalog",
+    "knowledge/patterns",
+    "templates/CLAUDE.md",
+    "templates/starter_profile.md",
+    "templates/temper_check.sh",
+    "templates/contrib_session.md",
+    "templates/goal-quench-hook-setup.md",
+    "templates/.claude/rules/session.md",
+    "scripts/below_floor_scan.sh",
+    "scripts/chamber_run.sh",
+    "scripts/fh_env_delta_scan.sh",
+    "scripts/substrate_jump_detector.sh",
+    "scripts/tier_census_grep.sh",
+    "scripts/test_marker_floor_lanes.sh",
+    "plugins/fh-meta/CHANGELOG.md",
     "knowledge/shared/learnings/subagent_invocations_log.yaml",
     "scripts/chamber_candidate_collect.sh",
     "scripts/fh_session_load.sh",
@@ -81,6 +102,12 @@
     "scripts/prepush_guard_check.sh",
     "scripts/psa_scan_lib.sh",
     "scripts/session_close_check.sh",
+    "scripts/memory_link_check.py",
+    "scripts/test_memory_link_check.sh",
+    "scripts/memory_nearcheck.py",
+    "scripts/sidecar_wait.sh",
+    "scripts/test_session_close_lanes.sh",
+    "scripts/test_card_drift_probe.sh",
     "scripts/universal_guard_check.sh",
     "scripts/test_prepush_stdin_integrity.sh",
     "scripts/public_surface_scan_files.sh",

package/plugins/fh-commons/.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "fh-commons",
-  "version": "1.4.73",
+  "version": "1.4.75",
   "engines": {
     "claudeCode": ">=1.0.0"
   },

package/plugins/fh-meta/.claude-plugin/plugin.json

@@ -1,10 +1,10 @@
 {
   "name": "fh-meta",
-  "version": "1.4.73",
+  "version": "1.4.75",
   "engines": {
     "claudeCode": ">=1.0.0"
   },
-  "description": "New in 1.4.73: the mechanical floor now ships — templates/.git-hooks (pre-commit/pre-push) plus every script they actually invoke, their regression anchors, and the .claude/rules data they read. Until now the package shipped the prose that tells you to run the gates while omitting the gates, and 35 shipped-doc references pointed at files the tarball did not contain. Verified by unpacking the tarball, installing the hooks in a scratch repo and committing through them, not by inspecting files[]. degrade_direction_scan gains shell probes: it collected .sh files that no probe could match and reported them clean (known-positive scored 0/4); git hooks were invisible entirely because the extension test ran against the path, not the basename. The field-propagated templates/ copy had drifted from scripts/ and is now byte-pinned by an anchor. Hub meta-engineering toolkit — 35 skills + 7 agents. New in 1.4.71: SKILL_detail.md brought inside the 4-axis gate — the gate matched the literal `SKILL.md`, which `SKILL_detail.md` does not contain, leaving 27.7% of the skill-spec surface ungated (measured; it leaked twice for real). Fix escalated from a name list to a directory scope after an enumeration sweep found a real uncovered file on its first run; anchored by scripts/gate_pathspec_check.sh (known-pair, wired into pre-commit). One-way mirror sync gains a destination-newer abort + mechanically injected mirror banner after two session cards were silently overwritten, anchored by scripts/sync_guard_check.sh. dialogue-harvest gains Step 4-b cross-corpus provenance (absorbed / held-unused / declined) for single-author corpora, where the original provenance labels were free and measured nothing. frontier-digest arxiv leg category-scoped after a relevance drift the staleness rule structurally could not catch (known-pair calibration recorded). templates/CLAUDE.md hub paths switched to {FH_ROOT} so a copied harness stops carrying one operator's absolute paths into org-visible checkouts. New in 1.4.70: dialogue-harvest (mines AI-dialogue logs — sycophancy strip first, induced-vs-independent provenance labeling, EN+KO known-pair calibration shipped); asset-placement-gate Step 0.6 official-corpora check (Claude Cookbook as Tier-0 consult corpus, judged-flag semantics); context-doctor built-in-/doctor-first sister anchor. New in 1.4.48: phantom-quench + steel-quench gain external frontier anchors (arXiv:2607.02052 package-hallucination; arXiv:2607.02057 prompt-coverage-adequacy); README model-flat claim reframed from a per-release point-curve to structural invariants (operation flattens across tiers; depth tier-order fixed within a generation). New in 1.4.47: onboarding step ① surfaces the Mode D companion-store session-start load in the auto-read salience anchor (previously only in the local binding + rules, so a greeting could skip the load). New in 1.4.46: context-doctor gains a command-output axis — routes to a command-output proxy/hook (rtk) to trim verbose CLI stdout, complementing .claudeignore; risk-gated to token-scarce environments (lossy filtering, off gate-input paths). New in 1.4.41: context-doctor 2026 trigger vocab (context engineering/rot/collapse) + phantom-citation hardening; hub measurement-integrity-checklist (cross-model measurement pre-flight: display-name pin/reps≥3/discriminating probe). New in 1.4.40: install-wizard scaffolds the companion store as a queryable wiki (INDEX + session-start read + Raw/Wiki/Conversation ingest axis). New in 1.4.39: auto-decorrelation (cross-family verifier sidecar recruitment, calibration-gated) + video-ingest (capability-routed video ingestion). New in 1.4.37: corpus-grounding-expander + persona-roster-expander (field-harvested verbatim-relay capability skills). New in 1.3.0: public-surface-audit (git-tracked private-token leak scan), field-harvest Mode B session-end auto-trigger, 4-axis gate scope extension (docs/ + AGENTS.md). New in 1.2.0: pipeline-conductor (4-pipeline gated sweep), return-path-gate (chain closure audit), goal-quench (Stop hook + quality gate), steel-quench Wave 5 (multi-model sidecar challenger), 2-layer architecture docs, YAML validation script. Validated cross-CLI: Claude Code, Codex, Gemini.",
+  "description": "New in 1.4.74: shipped documents no longer point at files the package omits. 35 such paths were measured on 2026-07-28 — CLAUDE.md instructed consumers to run templates/predelete_check.sh before a destructive op and the file did not ship. 20 more paths added; the 4 that remain absent are enumerated with a written reason each (per-environment registries and operator-private sync). Closing the set once is worth little because it regrows, so scripts/package_coverage_check.sh now fails selfcheck on any new occurrence, known-pair calibrated, and it caught itself on first wiring. New in 1.4.73: the mechanical floor now ships — templates/.git-hooks (pre-commit/pre-push) plus every script they actually invoke, their regression anchors, and the .claude/rules data they read. Until now the package shipped the prose that tells you to run the gates while omitting the gates, and 35 shipped-doc references pointed at files the tarball did not contain. Verified by unpacking the tarball, installing the hooks in a scratch repo and committing through them, not by inspecting files[]. degrade_direction_scan gains shell probes: it collected .sh files that no probe could match and reported them clean (known-positive scored 0/4); git hooks were invisible entirely because the extension test ran against the path, not the basename. The field-propagated templates/ copy had drifted from scripts/ and is now byte-pinned by an anchor. Hub meta-engineering toolkit — 35 skills + 7 agents. New in 1.4.71: SKILL_detail.md brought inside the 4-axis gate — the gate matched the literal `SKILL.md`, which `SKILL_detail.md` does not contain, leaving 27.7% of the skill-spec surface ungated (measured; it leaked twice for real). Fix escalated from a name list to a directory scope after an enumeration sweep found a real uncovered file on its first run; anchored by scripts/gate_pathspec_check.sh (known-pair, wired into pre-commit). One-way mirror sync gains a destination-newer abort + mechanically injected mirror banner after two session cards were silently overwritten, anchored by scripts/sync_guard_check.sh. dialogue-harvest gains Step 4-b cross-corpus provenance (absorbed / held-unused / declined) for single-author corpora, where the original provenance labels were free and measured nothing. frontier-digest arxiv leg category-scoped after a relevance drift the staleness rule structurally could not catch (known-pair calibration recorded). templates/CLAUDE.md hub paths switched to {FH_ROOT} so a copied harness stops carrying one operator's absolute paths into org-visible checkouts. New in 1.4.70: dialogue-harvest (mines AI-dialogue logs — sycophancy strip first, induced-vs-independent provenance labeling, EN+KO known-pair calibration shipped); asset-placement-gate Step 0.6 official-corpora check (Claude Cookbook as Tier-0 consult corpus, judged-flag semantics); context-doctor built-in-/doctor-first sister anchor. New in 1.4.48: phantom-quench + steel-quench gain external frontier anchors (arXiv:2607.02052 package-hallucination; arXiv:2607.02057 prompt-coverage-adequacy); README model-flat claim reframed from a per-release point-curve to structural invariants (operation flattens across tiers; depth tier-order fixed within a generation). New in 1.4.47: onboarding step ① surfaces the Mode D companion-store session-start load in the auto-read salience anchor (previously only in the local binding + rules, so a greeting could skip the load). New in 1.4.46: context-doctor gains a command-output axis — routes to a command-output proxy/hook (rtk) to trim verbose CLI stdout, complementing .claudeignore; risk-gated to token-scarce environments (lossy filtering, off gate-input paths). New in 1.4.41: context-doctor 2026 trigger vocab (context engineering/rot/collapse) + phantom-citation hardening; hub measurement-integrity-checklist (cross-model measurement pre-flight: display-name pin/reps≥3/discriminating probe). New in 1.4.40: install-wizard scaffolds the companion store as a queryable wiki (INDEX + session-start read + Raw/Wiki/Conversation ingest axis). New in 1.4.39: auto-decorrelation (cross-family verifier sidecar recruitment, calibration-gated) + video-ingest (capability-routed video ingestion). New in 1.4.37: corpus-grounding-expander + persona-roster-expander (field-harvested verbatim-relay capability skills). New in 1.3.0: public-surface-audit (git-tracked private-token leak scan), field-harvest Mode B session-end auto-trigger, 4-axis gate scope extension (docs/ + AGENTS.md). New in 1.2.0: pipeline-conductor (4-pipeline gated sweep), return-path-gate (chain closure audit), goal-quench (Stop hook + quality gate), steel-quench Wave 5 (multi-model sidecar challenger), 2-layer architecture docs, YAML validation script. Validated cross-CLI: Claude Code, Codex, Gemini.",
   "author": {
     "name": "chrono-meta",
     "email": "chrono-meta@users.noreply.github.com"