@event4u/agent-config 8.10.0 → 8.11.0

package/.claude-plugin/marketplace.json

@@ -6,7 +6,7 @@
   },
   "metadata": {
     "description": "Bootstrap shim for the event4u agent-config suite. This plugin ships ONLY the dispatcher hooks plus one install-pointer skill — ALL content (skills, rules, commands) is installed via `npx -y @event4u/agent-config init`, which writes the file projection AND registers the same hooks in ~/.claude/settings.json. DEPRECATED as a content channel; kept so the marketplace listing stays discoverable and existing installs keep working hooks.",
-    "version": "8.10.0",
+    "version": "8.11.0",
     "keywords": [
       "agent-config",
       "skills",

package/CHANGELOG.md

@@ -94,6 +94,23 @@ One question — answer in [Discussions](https://github.com/event4u-app/agent-co
 This gates the read-only MCP discovery helper (`agent-config mcp:search`) and the
 deferred auto-install question (`ADR-086`).
 
+### Added — humanizer (AI-tell removal for deliverable prose) (`road-to-humanized-writing` + `road-to-humanizer-hardening`)
+
+- **`humanizer` skill + `/humanize` command** (gtm-marketing pack) — remove
+  AI-writing tells from deliverable prose (posts, articles, drafts) via a
+  draft→audit→final loop; scoped to deliverable text only, never chat replies
+  or repo docs.
+- **`detect_ai_tells.ts` detector** — deterministic hard-tell / weighted-cluster
+  / em-dash-density scan (EN + DE), with a hidden-unicode ingestion scan and a
+  ReDoS length bound on untrusted input.
+- **Behavior change — write-engine step 4b is default-on.** Ghostwriter /
+  `post-as` drafts now pass a humanize audit before delivery; opt out per run
+  with `--raw`. The mechanical detector pass runs when a Node runtime is
+  present and degrades to a prose-only audit otherwise — the new
+  `humanizer-runtime` doctor check reports which path an install gets.
+- **Disclosure footer stays inviolable** through the audit; quoted/secondhand
+  text is never rewritten.
+
 ### Added — `agents/` directory contract + gitignore hygiene (`road-to-agents-dir-and-gitignore-hygiene`)
 
 - **`docs/contracts/agents-layout.md`** — authoritative contract for every top-level entry in `agents/` (purpose, git policy, retention, consumer-scope); includes the User Inbox Workflow (`agents/tmp/` → `agents/tmp.old/`).
@@ -998,6 +1015,137 @@ our recommendation order, not its support status.
 > that forces a new era split (`# Era: 8.10.x`, etc.) — see
 > [`docs/contracts/CHANGELOG-conventions.md § Era splits`](docs/contracts/CHANGELOG-conventions.md).
 
+## [8.11.0](https://github.com/event4u-app/agent-config/compare/8.10.0...8.11.0) (2026-07-12)
+
+### Features
+
+* **council:** chairman billable dispatch + estimate row + render injection ([9a62bf1](https://github.com/event4u-app/agent-config/commit/9a62bf1044a872949e46ffcb4581df4966d04887))
+* **council:** wire the stance tally end-to-end (final-round contract + Vote Tally render) ([ba84f55](https://github.com/event4u-app/agent-config/commit/ba84f55585a858f3d7053e950369a89b50e6f81c))
+* **council:** encode the contested-design council decisions (auto-select + repair policy) ([5f10a51](https://github.com/event4u-app/agent-config/commit/5f10a51904605ddd68840e3fa70983686e16fcdb))
+* **council:** chairman/debate_gates/restate config blocks + contract doc ([0f09a95](https://github.com/event4u-app/agent-config/commit/0f09a9525c010040facb84c3af57ce59eabfe968))
+* **council:** Phase 3 anti-conformity directive (wired) + debate-gate detectors ([b07b722](https://github.com/event4u-app/agent-config/commit/b07b7229caf61c793203dcd7ebe707f5d7b45ac1))
+* **council:** Phase 2 chairman selection (default-off, host fallback) ([caa518f](https://github.com/event4u-app/agent-config/commit/caa518fe27357200fa26fa66d5a55426986deb5f))
+* **memory:** five protocol borrows with existing checkers as hooks (retrieval Phase 3) ([9405ad1](https://github.com/event4u-app/agent-config/commit/9405ad169785cc0a8da20b5c390436b47c38425e))
+* **lexical:** term-coverage squaring in BM25 score (retrieval Phase 1) ([2beff15](https://github.com/event4u-app/agent-config/commit/2beff157a6b983d0b414f8521c552ac49be606a2))
+* **council:** Phase 1 — option-level stance-tally foundation (default-off) ([6852c40](https://github.com/event4u-app/agent-config/commit/6852c4080f5ca40514150a0e52491ac751b330e1))
+* **council:** Phase 0 — claims hygiene + verdict falsifiability ([c65885e](https://github.com/event4u-app/agent-config/commit/c65885e4bf3220fd32dabec53b706e682d8dfb67))
+* **worktree:** optional env-bootstrap stand-up convention (harness P1) ([b80e197](https://github.com/event4u-app/agent-config/commit/b80e1972ffe349feb98535b3e22c4dd8e919c454))
+* **bench:** spend-gate + claim re-scoping to the fixture corpus (Phases 3-4) ([52e0beb](https://github.com/event4u-app/agent-config/commit/52e0beb1486b6dbcd320324be5c9547c18f761b1))
+* **doctor:** humanizer-runtime check + step-4b fallback proof (Phase 2) ([f615ec6](https://github.com/event4u-app/agent-config/commit/f615ec6bea86138fec39f06cab46c8427974bb7f))
+* **humanizer:** untrusted-input handling on the ingestion path (Phase 1) ([902f9a4](https://github.com/event4u-app/agent-config/commit/902f9a4d83f67422dd7f47f0f14ed236bf99874b))
+* **design-fidelity:** no-unrequested-filler line + live-app-judge candidate ([ca57cc5](https://github.com/event4u-app/agent-config/commit/ca57cc53ee817fecef06277da38df26e1f6fc6ff))
+* **design-review:** canonical 6-state checklist + serial-audit decision ([a80de67](https://github.com/event4u-app/agent-config/commit/a80de67f84b7905b56a8bde5815b392ccfcda550))
+* **design:** consolidate slop-trope canon into one cited catalog ([dfe7f42](https://github.com/event4u-app/agent-config/commit/dfe7f42e9b9c026710a24ea4d0077ef05da01863))
+* **orchestration:** form gate, dispatch_mode telemetry, live-app judge mode, read-only knowledge attach ([8bf5990](https://github.com/event4u-app/agent-config/commit/8bf5990cfac3c445dd7e4cc6035f7f92953f3c73))
+* **subagents:** fleet hardening — prompt-defense preamble, claim witness sweep, originality shingles, hook-resilience shim ([ea40a54](https://github.com/event4u-app/agent-config/commit/ea40a54e62653843d82a288349ef7b7dc04d62a6))
+* **bench:** humanizer paired eval — corpus to 20 pairs, claim backed ([28dda19](https://github.com/event4u-app/agent-config/commit/28dda1942964841604d8ef148938c36dc209f4b0))
+* **commands:** /humanize — on-demand AI-tell removal cluster ([eb545e5](https://github.com/event4u-app/agent-config/commit/eb545e58367e8b0c8c0f8d6446409b4e207604f0))
+* **contracts:** write-engine step 4b — humanize audit ([29a1b9a](https://github.com/event4u-app/agent-config/commit/29a1b9a7f9114929f3db10a7cc006969be16d35c))
+* **skills:** humanizer skill — two-phase, deliverable-scoped ([489a143](https://github.com/event4u-app/agent-config/commit/489a14325c09aa3faf58d9d75ff9375146d43d61))
+* **scripts:** deterministic AI-tell detector with fixture corpus ([ae48a99](https://github.com/event4u-app/agent-config/commit/ae48a99d9dd8d0a6943f063901bd893b383cde38))
+* **flow-learnings:** real org fleet run green → close the roadmap ([71953fe](https://github.com/event4u-app/agent-config/commit/71953fe807152bab6b1492b3d5a977615438ae1a))
+* **knowledge:** flip global sharing default ON per ADR-119 (council-validated) ([be3f603](https://github.com/event4u-app/agent-config/commit/be3f6032640e40715e3a7c1ec145fe2508707501))
+* **token-saving:** Phase-0 quality gate measured — thin projection FAILS (does not ship) ([b362fa6](https://github.com/event4u-app/agent-config/commit/b362fa65369dab7209d6b9a8db98e705a8ac41a8))
+* **roadmaps:** add road-to-opt-* cluster from whole-package optimization sweep ([9688082](https://github.com/event4u-app/agent-config/commit/9688082a606a60879dc92aa29f5f575d23bc5a07))
+* **golden-set:** complete the consumer-scope corpus (council-labelled) + close golden-set-coverage ([fc6c120](https://github.com/event4u-app/agent-config/commit/fc6c120bb777182140f7ffc8fe905560179e9bff))
+* **domain-soundness:** ratify the validated count + close the roadmap ([c93f8bf](https://github.com/event4u-app/agent-config/commit/c93f8bf8e3dda437732024b65983aa68a09e76d4))
+* **orchestration:** deterministic parallelizable-classifier corpus recall (+regression test) ([7d78284](https://github.com/event4u-app/agent-config/commit/7d78284c25dd94068a45250dd66b21b331ce53f0))
+* **subagent-value:** verify orchestration telemetry pipeline end-to-end (Phase 1 Step 1) ([a0158e5](https://github.com/event4u-app/agent-config/commit/a0158e5f55a86e65db1fc72508fd270e72f4f2d9))
+* **orchestration-scope:** pre-register the falsifiable dispatch claim (Phase 1) ([1449f6c](https://github.com/event4u-app/agent-config/commit/1449f6c5bcb8d5efca87779a676fd9f10baf5f7b))
+* **evals:** weekly live trigger-eval pass-rate rotation in cross-model canary (ADR-118 §4) ([209c3fd](https://github.com/event4u-app/agent-config/commit/209c3fdf99fccd1cf2a8fa18d6c0b72cb2115bd9))
+* **flow-learnings:** record the live two-host matrix run (HONEST-NULL) + resolve the blocker ([4ad8c51](https://github.com/event4u-app/agent-config/commit/4ad8c51334fde4d54f09f4cfb39155e681aad4ed))
+* **release-ops:** leg manifest, bus-factor NEUTRAL state, growth ratchet, Primary-Goal ([0e70da1](https://github.com/event4u-app/agent-config/commit/0e70da15419b2a6d81f9f8aca9295448ea33ff91))
+* **worktree:** deterministic cleanup gates + 8-row edge-case matrix ([884c230](https://github.com/event4u-app/agent-config/commit/884c230f870be07b81a5866c85b4077e000b1b2e))
+* **evals:** priority-tier eval floor + two golden E2E evals ([7711ffe](https://github.com/event4u-app/agent-config/commit/7711ffe098ad3edce826a0e72f359619faadeb16))
+* **invariants:** documented change process for protected kernel strings ([dfb0e61](https://github.com/event4u-app/agent-config/commit/dfb0e6107d9f88246dcccd7716ffa2e98e9450be))
+* **rules:** code-comment-discipline — why-only comments, no signature-mirroring docblocks ([055a565](https://github.com/event4u-app/agent-config/commit/055a565a622536a318c8c0de3c3b2869ebfb198c))
+* **discipline-profile:** dispose of the full tier via council; park roadmap in later/ ([c03b21a](https://github.com/event4u-app/agent-config/commit/c03b21a85086fcb403d3e52b8147b5a56c4cea3f))
+
+### Bug Fixes
+
+* **hashes:** recondense the 7 commands whose assembled output my skill edits changed ([baa2afd](https://github.com/event4u-app/agent-config/commit/baa2afdd7b5641d6808fd65ba9342f0fba4300be))
+* **proof:** regenerate docs/proof.md after CLAIMS skill-count edit ([7e82690](https://github.com/event4u-app/agent-config/commit/7e826902a8fd47617a992e01be43b3947f2c3618))
+* **ai-council:** extract anthropic/gemini CLI default model constants ([850f465](https://github.com/event4u-app/agent-config/commit/850f4652b9caada116399bcc97ad67baaa5fec13))
+* **ai-council:** pin codex CLI default model as a named constant ([606f9ef](https://github.com/event4u-app/agent-config/commit/606f9ef45b44fe47c73f78105806bdba6df4ea00))
+* **index:** regenerate index.md + catalog.md after humanizer/humanize landed ([993ef76](https://github.com/event4u-app/agent-config/commit/993ef7626729b9989d8febf782578cd18f75e93d))
+* **roadmaps:** break `x` skill ref pattern for the not-yet-built evaluate-llm-feature ([245f02b](https://github.com/event4u-app/agent-config/commit/245f02b26c3c03474f42bad8b07b6cce2ccc7c68))
+* **contexts:** correct decision-revisit-gate link in the telegraph-tension note ([2711db0](https://github.com/event4u-app/agent-config/commit/2711db0acd2131f8df562fe31dacaeebc9c3802a))
+* **ci:** classify /humanize in surface-map flow coverage ([aef3224](https://github.com/event4u-app/agent-config/commit/aef322460a0e0b45b604c5accf55a218790c9d49))
+* **discovery:** skip _-prefixed subagent partials in strict manifest build ([dd943d6](https://github.com/event4u-app/agent-config/commit/dd943d64a73df659bd985e0e807181397ab416e6))
+* **ci:** humanize command visibility internal (tier-2 contract) ([b715105](https://github.com/event4u-app/agent-config/commit/b71510564fcf6f8e3754f84356a9786cd84e164a))
+* **ci:** eslint no-console + unused import; humanize visibility enum ([735eb03](https://github.com/event4u-app/agent-config/commit/735eb034c3e626d2d53bad440c881e4660d2a1f3))
+* **legal:** reflow forbidden-language examples onto marker-carrying lines ([d3496b9](https://github.com/event4u-app/agent-config/commit/d3496b91805d16ae07def730669182eeeb15ba25))
+* **neutrality:** framework-leakage allowlist 104 -> 14 via linter upgrades + 25 neutralizations + 2 carve-outs ([0f04776](https://github.com/event4u-app/agent-config/commit/0f047762a28067f0422601e417bbb87af88b892c))
+* **install:** rebuild bundle with local node_modules (path-comment drift) ([ce3d845](https://github.com/event4u-app/agent-config/commit/ce3d8456e58d9e09da7c73282b538dc46fe91040))
+* **self-review-gate:** never fail CI on missing key / no credit — wrap the live path ([8005839](https://github.com/event4u-app/agent-config/commit/8005839fc1c7447fa91d80cf30b2ed29f66b0a0a))
+* **roadmaps:** close the two transcript-audit gaps ([e3bc430](https://github.com/event4u-app/agent-config/commit/e3bc430aba27fe88b70db0a016a4e19bac9b2e92))
+* **roadmaps:** review nits — benchmark scope precision, SHA-rule semantics, resolve verified borrow ([b0ed571](https://github.com/event4u-app/agent-config/commit/b0ed5712e1ef92b7e4084d657dba5ebb2236ba3d))
+* **roadmaps:** review corrections — SHA-pinned verification, drop landed scope, fix numbers ([14025ce](https://github.com/event4u-app/agent-config/commit/14025cef2c6aee1a058f7ad3048f580cd9ce667f))
+* **install:** ship the missing dist/install/rule_scope.js (stop-hook crash) ([4222081](https://github.com/event4u-app/agent-config/commit/422208190e9bc4a56cc82baa795127224e8c1b65))
+* **rules:** wire code-comment-discipline into router, counts, and settings schema ([ffa88e5](https://github.com/event4u-app/agent-config/commit/ffa88e546335b1832a82cc7cb9067845b09bb441))
+* **proof:** compute the coverage paragraph live from computeCoverage() ([dfa77e9](https://github.com/event4u-app/agent-config/commit/dfa77e9c6e180d21be0b86afa555f53c8f79619d))
+* **install:** rebuild install bundle with worktree-local module paths ([5933bf1](https://github.com/event4u-app/agent-config/commit/5933bf1a9b9b46baf5b38f860bdee020c8b4a2dc))
+
+### Documentation
+
+* **roadmap:** flip the chairman-dispatch + estimate-row steps ([a2faf9d](https://github.com/event4u-app/agent-config/commit/a2faf9dc576a14c984ffc23daf9c083166254a0c))
+* **roadmap:** flip the stance-tally wiring steps (Phase 1 integration live) ([3117878](https://github.com/event4u-app/agent-config/commit/31178781106a3fc0a31c3710ee42e9b7037b50ad))
+* **roadmap:** fix broken ref — name upstream plugin files without local-path form ([7dec6df](https://github.com/event4u-app/agent-config/commit/7dec6dfdabefd4abe9edcf4244215a3827dcb2de))
+* **roadmap:** resolve contested-design-council-pass with council notes ([1fd6dff](https://github.com/event4u-app/agent-config/commit/1fd6dff6f1f4f3a2080ae3c1531515bee231f34d))
+* **roadmap:** add team-mode roadmap (govern the official cross-model pair) ([a185db0](https://github.com/event4u-app/agent-config/commit/a185db0de5440923cc4045cb49f25fd29dbe832d))
+* **ai-council:** document council vs team-mode boundary ([af7ffeb](https://github.com/event4u-app/agent-config/commit/af7ffeb9bf32a06dea1ee5439fc9f8d9cfca363e))
+* **readme:** drop artefact counts from prose — badges carry the numbers ([4ea0f47](https://github.com/event4u-app/agent-config/commit/4ea0f47c5be35554d8c40b214492bfe7ccff93f8))
+* **design:** measurement-unblock design prep + un-park (runs gated) ([72f618a](https://github.com/event4u-app/agent-config/commit/72f618acbadd61a56fa65b0939bf3c87d57f3269))
+* **roadmaps:** reconcile parallel-exploration depth into the harvest cluster ([aa50574](https://github.com/event4u-app/agent-config/commit/aa505749a625b0be321982313e435cfbfe27b0c3))
+* **roadmaps:** second-sweep harvest — review, workflow, quality-gates + coverage ([d2258ed](https://github.com/event4u-app/agent-config/commit/d2258ed5dc61c796091eaa8751ba92f500f4804a))
+* **roadmaps:** add ecosystem-harvest roadmap set from external skill survey ([d5bed31](https://github.com/event4u-app/agent-config/commit/d5bed315a458ee9e8ba98af4a4ec62308e069d88))
+* **roadmaps:** close + archive humanizer-hardening (21/21) ([1ad6813](https://github.com/event4u-app/agent-config/commit/1ad681324ac25109c8590ca7b0b45090ff91ef55))
+* **roadmaps:** humanizer-hardening follow-up from parent review ([f69621e](https://github.com/event4u-app/agent-config/commit/f69621e6ebbb665da15420ed3c8567f8c1c004f5))
+* **roadmaps:** humanized-writing adoption roadmap (council-decided) ([464a66c](https://github.com/event4u-app/agent-config/commit/464a66c0ae48abc82a951ea8201580f6eab27474))
+* **roadmap:** road-to-loop-engineering executed and archived ([ccad788](https://github.com/event4u-app/agent-config/commit/ccad788bb78b84346abf811119eecf923cb21501))
+* **adr:** ADR-118 loop-engineering boundaries — one closure, four rejections, zero new loop surfaces ([c27ba0c](https://github.com/event4u-app/agent-config/commit/c27ba0cee1295f7a88e231e1a458c35efba13c70))
+* **telemetry:** operationalize first_pass_success / escalated — zero schema change ([dfc8a64](https://github.com/event4u-app/agent-config/commit/dfc8a646a62b1cc03876d53f687d2673dff44d25))
+
+### Refactoring
+
+* **rules:** thin-stub the five heaviest auto-rules (-5,652 eager tokens) ([6ef4102](https://github.com/event4u-app/agent-config/commit/6ef4102d6448cf5a37f3e1555b33299ebd75c907))
+
+### Tests
+
+* **humanizer:** lock scan-what-you-ingest invariant + spend-gate no-network ([117a703](https://github.com/event4u-app/agent-config/commit/117a7039b6ad6496d787ac44d9d61f02f7b35351))
+* **subagents:** exclude _-prefixed partials from wedge coverage ([f117d38](https://github.com/event4u-app/agent-config/commit/f117d3893c2ae83219d202c4139933a71b539bcb))
+
+### Chores
+
+* **condense:** refresh folded condensation hashes for 11 dependents ([701d588](https://github.com/event4u-app/agent-config/commit/701d588962f1ee807689820318e9a6fe82988892))
+* **skills:** clear three chronic skill-lint warnings ([c6c62af](https://github.com/event4u-app/agent-config/commit/c6c62af0162a723fbc083cbe3e1979b1ff2fa3a6))
+* **roadmap:** council-deliberation Phase 2-3 continuation (default-off, dispatch gated) ([197ce1b](https://github.com/event4u-app/agent-config/commit/197ce1b368b5ab11b0b55d1a38cddf5d70561ac7))
+* ignore fleet ([6b7a381](https://github.com/event4u-app/agent-config/commit/6b7a3816136541f6f31e59f25d465b74154f86dd))
+* **roadmap:** un-park retrieval-and-memory (Phases 1/3/4 + Phase-2 κ) + hashes ([979de16](https://github.com/event4u-app/agent-config/commit/979de160ef90c062d994ba0398adfa9eff345da6))
+* **roadmap:** un-park council-deliberation (Phase 0 done, Phase 1 foundation) + hashes ([a56bee6](https://github.com/event4u-app/agent-config/commit/a56bee6e4b507c6271f69bde2ca2842e6483309b))
+* **roadmap:** resolve harness-discipline P5 (route to council) + archive ([83952cf](https://github.com/event4u-app/agent-config/commit/83952cfdffeb69f9d995a4fc373c4e8c4bad5073))
+* **roadmap:** un-park harness-discipline (Phases 1-4 done, P5 open) + hashes ([16127fa](https://github.com/event4u-app/agent-config/commit/16127fa1883786c676dda3b7934c32aceb608f96))
+* **roadmap:** archive road-to-opt-design-polish + refresh condensation hashes ([c57932d](https://github.com/event4u-app/agent-config/commit/c57932dc266ea67ffff0b38fa1e95df8a96699b1))
+* regenerate wedge doc + meta catalog after prompt-defense/orchestration edits ([28b87fd](https://github.com/event4u-app/agent-config/commit/28b87fdda66e0447dcc52be4e4737381d4a4490b))
+* refresh condensation hashes for subagent-orchestration dependents ([6c1f61d](https://github.com/event4u-app/agent-config/commit/6c1f61d5639d904a512aac87f00f272f8acef715))
+* **proof:** regenerate docs/proof.md after CLAIMS.md humanizer entry ([9c6099e](https://github.com/event4u-app/agent-config/commit/9c6099e4627e3d46ec0af7c3025051af168552fd))
+* **roadmaps:** archive completed humanized-writing roadmap ([6892409](https://github.com/event4u-app/agent-config/commit/6892409d27e9346f3bcaf90b29b7ca46fc31748c))
+* **hygiene:** reports retention convention, reference justification, orphan sweep, roadmap closure ([eb2b98e](https://github.com/event4u-app/agent-config/commit/eb2b98e0b1ccfd965efa073f640e8ec125946b1b))
+* **roadmaps:** AI-council re-scope of the token/measurement cluster after the thin-null ([74120fd](https://github.com/event4u-app/agent-config/commit/74120fd15ed549f18494d2cddd529200f0653aa6))
+* **adr:** decision-flips hygiene — ADR-092 addendum, corpus repairs, stale-prose fixes ([e405190](https://github.com/event4u-app/agent-config/commit/e405190dc0620bb7f859c442033d31a082ac6204))
+* **token-saving:** reality-sync 2 verified-met acceptance criteria (Phase 5 + Phase 10) ([faea1de](https://github.com/event4u-app/agent-config/commit/faea1ded3e7bdc81697a562ad7bcbe0d194160c9))
+* **roadmaps:** execute road-to-opt-portfolio-consolidation (process-full) ([c54ea39](https://github.com/event4u-app/agent-config/commit/c54ea39984dc436e14be0a743e60f7d82b435687))
+* **request-scoped:** record the held-quality gate as evidence-RED (thin failed), no further spend ([7642c13](https://github.com/event4u-app/agent-config/commit/7642c131ba3c70b20ae0c24447f911f4dc8b8149))
+* **roadmaps:** park the road-to-opt-* cluster in later/ (maintainer decision) ([9a027fe](https://github.com/event4u-app/agent-config/commit/9a027fe8cdd7d68dfc3636ccac652771adc9b463))
+* modify domain soundness ([b51eb29](https://github.com/event4u-app/agent-config/commit/b51eb29dbaa4d3ad2c7af13c32df8cfa0347c276))
+* **gitignore:** re-ignore dist/install/*_cli.js tsc byproducts ([e5913ef](https://github.com/event4u-app/agent-config/commit/e5913ef0404ee99c9c337b404882616f25edb763))
+* **packs:** regenerate engineering-base manifest for code-comment-discipline ([3973c47](https://github.com/event4u-app/agent-config/commit/3973c47651a60a617363786a13b899b752a01f1c))
+* **proof:** regenerate proof.md for the new unbacked claim (build-proof-check) ([7cbe82f](https://github.com/event4u-app/agent-config/commit/7cbe82fa828335e78b33cbb7b7c94c7472434440))
+* **roadmap:** fable-feedback-5 — executed and archived in its own PR ([e106f40](https://github.com/event4u-app/agent-config/commit/e106f40dfe0174f1678f7b8f5a13a2d3859b6d53))
+* **flow-learnings:** commit the bench matrix run-config + concretize the live-run blocker ([318739c](https://github.com/event4u-app/agent-config/commit/318739c0ef10a79efe899052780340a646797567))
+
+Tests: 7379 (+167 since 8.10.0)
+
 ## [8.10.0](https://github.com/event4u-app/agent-config/compare/8.9.0...8.10.0) (2026-07-10)
 
 ### Features

package/README.md

@@ -4,7 +4,7 @@
 
 [![Smoke](https://github.com/event4u-app/agent-config/actions/workflows/smoke.yml/badge.svg)](https://github.com/event4u-app/agent-config/actions/workflows/smoke.yml) [![Public install smoke (3 OS × 2 Node)](https://github.com/event4u-app/agent-config/actions/workflows/smoke-public-install.yml/badge.svg)](https://github.com/event4u-app/agent-config/actions/workflows/smoke-public-install.yml) [![npm](https://img.shields.io/npm/v/@event4u/agent-config?style=flat-square&label=npm&color=orange)](https://www.npmjs.com/package/@event4u/agent-config) [![agent-config MCP server](https://glama.ai/mcp/servers/event4u-app/agent-config/badges/score.svg)](https://glama.ai/mcp/servers/event4u-app/agent-config)
 
-[![Skills](https://img.shields.io/badge/Skills-270-orange?style=flat-square)](dist/agent-src/skills/) [![Rules](https://img.shields.io/badge/Rules-103-orange?style=flat-square)](dist/agent-src/rules/) [![Commands](https://img.shields.io/badge/Commands-177-orange?style=flat-square)](dist/agent-src/commands/) [![Guidelines](https://img.shields.io/badge/Guidelines-90-orange?style=flat-square)](docs/guidelines/) [![Personas](https://img.shields.io/badge/Personas-29-orange?style=flat-square)](dist/agent-src/personas/) [![Advisors](https://img.shields.io/badge/Advisors-5-orange?style=flat-square)](dist/agent-src/personas/advisors/)
+[![Skills](https://img.shields.io/badge/Skills-271-orange?style=flat-square)](dist/agent-src/skills/) [![Rules](https://img.shields.io/badge/Rules-104-orange?style=flat-square)](dist/agent-src/rules/) [![Commands](https://img.shields.io/badge/Commands-178-orange?style=flat-square)](dist/agent-src/commands/) [![Guidelines](https://img.shields.io/badge/Guidelines-90-orange?style=flat-square)](docs/guidelines/) [![Personas](https://img.shields.io/badge/Personas-29-orange?style=flat-square)](dist/agent-src/personas/) [![Advisors](https://img.shields.io/badge/Advisors-5-orange?style=flat-square)](dist/agent-src/personas/advisors/)
 
 > **Choose your experience — developer · founder · content · agency · finance · ops. Add packs. Get a focused command set, not a 500-artefact dump.** Bring your own AI provider.
 
@@ -12,13 +12,13 @@
 
 **Every public claim in this README is machine-checked — [verify it yourself](docs/proof.md).** In a market that runs on unbacked headline numbers, this one binds each claim to resolvable evidence or fails its own build.
 
-**270 skills, 177 commands, 103 governed rules** — plus a capability router that loads the right skill on intent, and multi-agent orchestration with consensus review. The whole layer is compiled into 7+ host agents (Claude Code, Cursor, Augment, Cline, Windsurf, Copilot, Gemini) with **zero runtime daemon**.<!-- claim:no-runtime-daemon --> Six role-shaped entry paths sit on top, so any host becomes a reliable team member — without locking you to a single model or vendor.
+**A deep library of skills, commands, and governed rules** — plus a capability router that loads the right skill on intent, and multi-agent orchestration with consensus review. The whole layer is compiled into 7+ host agents (Claude Code, Cursor, Augment, Cline, Windsurf, Copilot, Gemini) with **zero runtime daemon**.<!-- claim:no-runtime-daemon --> Six role-shaped entry paths sit on top, so any host becomes a reliable team member — without locking you to a single model or vendor.
 
 ### What's different
 
 It is both deep **and** disciplined — and honest about what it deliberately is not:
 
-- **Depth that routes itself** — 270 skills + 177 commands, with a capability router that loads the right one on intent, not a 500-artefact context dump.
+- **Depth that routes itself** — a deep skill and command library, with a capability router that loads the right one on intent, not a 500-artefact context dump.
 - **Governance on every host** — rules compiled into each tool's native format at projection time; deterministic runtime hooks added on hook-capable hosts. This config-space, host-agnostic governance is the moat ([the governance advantage](docs/governance-advantage.md) · [enforcement by host](docs/enforcement-by-host.md)).
 - **Surgical uninstall** — removes only its own keys from a shared host config (matched by JSON-pointer + SHA-256), never a neighbour tool's entries.<!-- claim:surgical-uninstall -->
 - **Pack-scoped install** — writes the active pack only, not a 500-artefact dump.
@@ -67,7 +67,7 @@ what is *not* loaded · examples):
 
 ### Workflows, not raw commands
 
-You don't memorize 177 commands — you run a **work journey**. Four flows span the
+You don't memorize every command — you run a **work journey**. Four flows span the
 developer story end-to-end; each names the command you TYPE to start and the
 skills it composes:
 
@@ -462,7 +462,7 @@ When a prompt matches a command's purpose ("setze ticket ABC-123 um" → `/imple
 | [**Migration**](docs/MIGRATION.md) | Per-version upgrade steps |
 | [**Showcase**](docs/showcase.md) | More examples & expected behavior |
 
-Browse content: [all 177 commands](dist/agent-src/commands/) · [skills catalog](docs/skills-catalog.md) · [full catalog](docs/catalog.md) · [`llms.txt`](llms.txt).
+Browse content: [all commands](dist/agent-src/commands/) · [skills catalog](docs/skills-catalog.md) · [full catalog](docs/catalog.md) · [`llms.txt`](llms.txt).
 
 ---
 

package/dist/agent-src/commands/challenge-me/vision.md

@@ -34,7 +34,10 @@ Drop the seed — a plan, idea, ticket draft, or design sketch — and I'll
 walk you through a decision tree, one question per turn, until I'm 95%
 sure what you want. Then you get a copyable Markdown pitch.
 
-Stop early any time by typing `!pitch` (or `!Vision pitchen`).
+End the interview early any time:
+  `!pitch`              — emit pitch, hand back
+  `!roadmap`            — emit pitch + auto-route to `/roadmap:create`
+  `!ai` / `!council`    — poll the AI council for the open question(s), continue (cost auto-accepted)
 ```
 
 Skip the welcome if the user invokes `/challenge-me vision <seed>` directly.
@@ -138,7 +141,7 @@ NEVER ASK WHAT `grep`, `view`, OR `codebase-retrieval` ANSWERS IN SECONDS.
 2. **Look up before asking.** Existing routes, model fields, conventions,
    feature flags, previous similar features → resolve from the repo.
    Surface findings briefly in the next question block ("Repo says X
-   already exists at `path/to/file.php`"). The user should never have to
+   already exists at `path/to/file`"). The user should never have to
    answer a question the codebase already answers.
 3. **List the open dimensions** internally — goal, scope, users, data,
    constraints, success criteria, edge cases, rollback. Do not surface
@@ -164,9 +167,11 @@ constraint this unblocks>.
 3. <alternative option>
 4. Skip / not relevant
 
-> Reply with a number, a free-form answer, or type **`!pitch`**
-> (or **`!Vision pitchen`**) to lock in the current understanding and
-> get the Markdown pitch.
+> Reply with a number, a free-form answer, or one of:
+> **`!pitch`** — emit pitch and hand back ·
+> **`!roadmap`** (or **`!roadmap:create`**) — emit pitch + auto-route to `/roadmap:create` ·
+> **`!ai`** (alias **`!ai-council`** / **`!council`**) — poll the council on the open question(s),
+> cost auto-accepted, then continue the interview.
 ````
 
 **Rules for the question block.**
@@ -188,25 +193,37 @@ constraint this unblocks>.
 1. Number reply (`1`, `2`, …) → adopt that option.
 2. Free-form reply → integrate verbatim, do not re-interpret.
 3. "Skip" / "skip" / "n/a" → branch closed, do not revisit unless reopened.
-4. Pitch trigger fired → jump straight to Step 4, even if confidence < 95%.
+4. Trigger fired → branch by trigger type (see table below):
+   - `!pitch` → jump to Step 4, then Step 5 (emit pitch), then Step 6 hand-back.
+   - `!roadmap` / `!roadmap:create` → Step 4 → Step 5 (emit pitch) → Step 6 with **roadmap routing**.
+   - `!ai` / `!ai-council` / `!council` → jump to **Step 4-bis (Council poll)**, then resume Step 1 with the answers integrated.
 
-#### Pitch trigger — strict matching
+#### Triggers — strict matching
 
-The pitch trigger requires **explicit syntax** to avoid false positives
-on natural sentences like "I'd pitch this to the team".
+Three trigger families end or branch the interview. All require
+**explicit syntax** to avoid false positives on natural prose
+("I'd pitch this to the team", "let's run a council later").
+
+| Trigger | Action | Cost gate |
+|---|---|---|
+| `!pitch` · `/pitch` · whole reply `pitch` | Emit pitch (Step 5), hand back (Step 6) | none |
+| `!roadmap` · `!roadmap:create` · `/roadmap` · `/roadmap:create` · whole reply `roadmap` | Emit pitch (Step 5), then route to `/roadmap:create` with the pitch as seed (Step 6) | none — roadmap creation is text-only |
+| `!ai` · `!ai-council` · `!council` · `/ai` · `/ai-council` · `/council` · whole reply `ai` / `council` | Snapshot open question(s), invoke `/council default` to draft answers, integrate, resume Step 1 | **auto-accepted** — typing the trigger is the consent (per user opt-in to the trigger contract) |
+
+Match rules apply to **every** trigger above:
 
 | Match | Example | Fires? |
 |---|---|---|
-| `!pitch` (or any `!`-prefixed variant) at line start | `!pitch`, `!Vision pitchen`, `!pitch it` | yes |
-| `/pitch` at line start | `/pitch` | yes |
-| Whole reply is exactly `pitch` (case-insensitive, no other words) | `pitch` | yes |
-| `pitch` inside a sentence | "I would pitch this on Monday" | **no** |
-| `pitch mich` / `ship vision` inside prose | natural-language fragment | **no** |
+| `!<trigger>` at line start | `!pitch`, `!roadmap`, `!council`, `!Vision pitchen` | yes |
+| `/<trigger>` at line start | `/pitch`, `/roadmap:create`, `/ai-council` | yes |
+| Whole reply exactly the trigger word (case-insensitive) | `pitch`, `roadmap`, `council` | yes |
+| Trigger word inside a sentence | "I would pitch this on Monday", "let's poll the council later" | **no** |
+| Natural-language fragment (`pitch mich`, `ship vision`) | prose | **no** |
 
-`DE: "!Vision pitchen", "/Vision pitchen", "/pitche das", whole reply "pitch" · EN: "!pitch the vision", "/pitch the vision", whole reply "pitch"`
+`DE: "!Vision pitchen", "!roadmap erstellen", "!Council befragen", whole reply "pitch" · EN: "!pitch the vision", "!roadmap it", "!council on this", whole reply "pitch"`
 
-If the agent is unsure whether the user pitched, it confirms once
-(numbered options: `1. emit pitch · 2. continue interview`) instead of
+If the agent is unsure **which** trigger fired (e.g. `!council and pitch`),
+it confirms once with numbered options listing the candidates instead of
 guessing.
 
 ### Step 3: Re-score and continue
@@ -218,7 +235,7 @@ the next question. If reached, go to Step 4 without asking.
 
 Before emitting the pitch, run this checklist internally:
 
-1. All four 95%-conditions hold (or the user fired the pitch trigger).
+1. All four 95%-conditions hold (or the user fired `!pitch` / `!roadmap`).
 2. Every adopted answer is reflected in the pitch — no orphan branches.
 3. Acceptance criteria are observable and testable in the project's
    existing test surface — not "works correctly" / "looks better".
@@ -226,6 +243,41 @@ Before emitting the pitch, run this checklist internally:
 
 If any check fails, ask one more question instead of pitching.
 
+### Step 4-bis: Council poll (`!ai` / `!ai-council` / `!council` only)
+
+Fires **instead of** Step 4/5 when an AI trigger matches. Cost gate is
+auto-accepted by the trigger contract — do **not** re-ask.
+
+1. **Snapshot** the current interview state:
+   - The seed (verbatim).
+   - All adopted answers so far (as `<question> → <answer>` lines).
+   - Open branches and the very-next question on the table.
+   - Any open assumptions surfaced so far.
+2. **Build the council prompt** as a single block:
+
+   ```
+   You are reviewing an in-progress /challenge-me vision interview.
+   Seed: <verbatim>
+   Adopted: <list>
+   Open question(s): <list>
+   For each open question, recommend an answer with one-line rationale,
+   plus any unstated risk you spot. Be terse.
+   ```
+3. **Invoke** `/council default prompt:"<above>"` non-interactively
+   (auto-accepted cost gate). Pass the host's `ai_council.min_rounds`
+   default — do **not** raise rounds inside the trigger.
+4. **Integrate** the council output:
+   - Surface each council answer **labelled with the source** (e.g.
+     `Council recommends: <answer> — rationale: <one line>`).
+   - Do **not** silently adopt — the user still picks (`1` to accept,
+     free-form to refine, `skip` to discard).
+5. **Resume Step 1** with the very-next question, now showing the
+   council's recommendation as option 1 (with the existing recommended
+   option pushed to option 2 if they differ).
+
+If the council call fails (network, budget, missing member), report the
+error in one line and resume Step 1 unchanged — do **not** re-trigger.
+
 ### Step 5: Emit the pitch
 
 Output a single fenced Markdown block the user can copy verbatim. Shape:
@@ -262,22 +314,50 @@ Output a single fenced Markdown block the user can copy verbatim. Shape:
 The outer fence uses four backticks so the inner triple-backtick stays
 literal when the user copies it.
 
-### Step 6: Hand back
+### Step 6: Hand back — HARD STOP after artifact
 
-After the pitch is emitted, end the turn. Do **not** propose to also
-implement the pitch, write it to a roadmap, or commit anything — the
-pitch is conversational output the user copies into the next command
-(`/work`, `/roadmap-create`, `/feature-plan`, …).
+```
+PITCH OR ROADMAP LANDED → STOP. NEVER PROPOSE IMPLEMENTATION,
+NEVER ASK "READY TO START?". THE USER PICKS THE NEXT MOVE WITH AN
+EXPLICIT EXECUTION VERB ON A LATER TURN.
+```
 
-If the user wants the next step run, they invoke it explicitly.
+Branch on which trigger fired (or the natural 95%-stop):
+
+- **Natural stop or `!pitch`** — end the turn after emitting the pitch.
+  Do **not** propose to also implement, write to a roadmap, or commit —
+  the pitch is conversational output the user copies into the next
+  command (`/work`, `/roadmap:create`, `/feature-plan`, …).
+- **`!roadmap` / `!roadmap:create`** — emit the pitch (Step 5) **and**
+  immediately route to [`/roadmap:create`](../roadmap/create.md) using
+  the pitch as the seed for its Step 1 ("Title / goal", "Context",
+  "Phases"). The user keeps full control inside `/roadmap:create` —
+  this is opt-in routing, not silent execution. Mention the route in a
+  single line under the pitch (e.g. *"Routing to `/roadmap:create` —
+  this pitch becomes the seed for title + context + phase 1."*).
+  The chain ends when `/roadmap:create` saves the file — its own Step
+  9 is a hard stop ([`roadmap/create § 9`](../roadmap/create.md)), no
+  execution offer rides on the back of `!roadmap`.
+- **`!ai` / `!ai-council` / `!council`** — handled by Step 4-bis, never
+  reaches Step 6.
+
+If the user wants any other follow-up, they invoke it explicitly. The
+"Recommended next step" line in the pitch (Step 5) names a *command*
+(`/work`, `/roadmap:create`, …) — it is a suggestion the user copies,
+**not** an offer this command auto-routes through.
 
 ## Output format
 
 1. **Per-turn question block** — one numbered question with a recommended
-   option, until 95% or pitch trigger.
+   option, until 95% or a trigger (`!pitch` / `!roadmap` / `!ai`).
 2. **Final pitch** — a single copyable fenced Markdown block (Step 5 shape).
 3. **One-line handoff suggestion** under the pitch — which command to run
-   next (`/work`, `/roadmap-create`, `/feature-plan`, etc.).
+   next (`/work`, `/roadmap:create`, `/feature-plan`, etc.).
+   - For `!roadmap` triggers, this line is the route announcement
+     instead of a suggestion (Step 6).
+4. **Council-poll output** (only if `!ai` fired) — labelled council
+   answers integrated into the next question block (Step 4-bis), not a
+   final pitch. The interview resumes.
 
 ## Examples
 
@@ -310,17 +390,26 @@ If the user wants the next step run, they invoke it explicitly.
 - **One question per turn** — non-negotiable, per
   [`ask-when-uncertain`](../../rules/ask-when-uncertain.md) Iron Law.
 - **No auto-execution** — emit the pitch, not the implementation it
-  describes. Do not chain into `/work` or `/roadmap-create` without
-  explicit user invocation.
-- **No file writes** — the pitch is conversational. No commits, no
-  roadmap edits, no `agents/` writes.
+  describes. Do not chain into `/work` or `/roadmap:create` without
+  explicit user invocation. The `!roadmap` trigger **is** explicit
+  invocation by user opt-in; routing under it is permitted **and ends
+  at file save** — never propose "ready to start Phase 1?" after the
+  roadmap lands. See [`scope-mechanics § Post-artifact hard stop`](../../contexts/authority/scope-mechanics.md).
+- **No file writes from /challenge-me itself** — the pitch is
+  conversational. No commits, no roadmap edits, no `agents/` writes.
+  The `!roadmap` trigger routes to `/roadmap:create`, which writes
+  under its own contract (with its own user confirmation in Step 6).
+- **Council cost gate is auto-accepted under `!ai`** — typing the
+  trigger is the consent. Do not surface a second confirmation. If
+  `personal.autonomy: auto`, the trigger still consents on its own —
+  the user explicitly opted in by typing it.
 - **Codebase first** — Step 0 forbids asking what `grep`, `view`, or
   `codebase-retrieval` would answer in seconds.
 - **Welcome once** — render the welcome only on the first no-body
   invocation per session.
 - **Mirror the user's language** — the question blocks and pitch use
-  the language the user wrote in (`language-and-tone` Iron Law). The
-  pitch trigger syntax (`!pitch`, `/pitch`, whole reply `pitch`) stays
+  the language the user wrote in (`language-and-tone` Iron Law). All
+  trigger syntax (`!pitch`, `!roadmap`, `!ai` and their aliases) stays
   literal in any language.
 
 ## Do NOT

package/dist/agent-src/commands/challenge-me/with-docs.md

@@ -22,8 +22,10 @@ packs:
 > Same one-question-at-a-time interview as `/challenge-me vision`, plus:
 > a **session glossary** against `CONTEXT.md`, **load-bearing claim-vs-code
 > verification**, and an optional `CONTEXT.md` **patch** + **ADR candidates**
-> appended to the final pitch. Nothing is written to disk — the pitch is
-> the only artefact.
+> appended to the final pitch. `/challenge-me` itself never writes to
+> disk — the in-interview triggers (`!roadmap`, `!ai`) route to
+> `/roadmap:create` and `/council default`, which run under their own
+> contracts.
 
 ## Welcome
 
@@ -36,7 +38,10 @@ CONTEXT.md and verify load-bearing claims against the code. At the end
 you get a pitch — plus an optional CONTEXT.md patch and ADR candidates
 to copy / apply yourself.
 
-Stop early any time by typing `!pitch` (or `!Vision pitchen`).
+End the interview early any time:
+  `!pitch`              — emit pitch, hand back
+  `!roadmap`            — emit pitch + auto-route to `/roadmap:create`
+  `!ai` / `!council`    — poll the AI council for the open question(s), continue (cost auto-accepted)
 ```
 
 Skip the welcome if the user invokes `/challenge-me with-docs <seed>`
@@ -56,18 +61,26 @@ If neither file exists and the user invoked `with-docs` anyway, fall
 back gracefully to `vision`-style flow and skip the doc inventory; note
 the absence in a single line at the end of Step 0.
 
-## Stop condition, flags, pitch trigger
+## Stop condition, flags, triggers
 
 Inherits **verbatim** from [`vision.md`](vision.md):
 
 - Default stop = the four 95%-conditions (AND, not OR).
 - Load-bearing test = answer changes goal / scope / hard constraint / AC.
 - `--until=95%` (default) · `--until=N` · `--keep-going`.
-- Pitch trigger: `!pitch`, `/pitch`, or whole reply `pitch`.
-- Step 1 question-block shape, Step 4 pitch validation.
+- Triggers (full syntax + match rules in `vision.md` § Triggers):
+  - `!pitch` / `/pitch` / whole reply `pitch` — emit pitch, hand back.
+  - `!roadmap` / `!roadmap:create` (and `/`-prefixed) — emit pitch + auto-route to `/roadmap:create` with the pitch as seed.
+  - `!ai` / `!ai-council` / `!council` (and `/`-prefixed) — invoke `/council default` on the open question(s), cost auto-accepted, integrate, resume.
+- Step 1 question-block shape, Step 4 pitch validation, Step 4-bis council poll.
 
 The deltas below replace / extend Step 0, Step 1, Step 2, Step 4 and
-Step 5. Steps 3 and 6 are unchanged.
+Step 5. Steps 3 and 6 are unchanged. **`!roadmap` routing in Step 6**
+includes the optional `CONTEXT.md` patch + ADR candidates from this
+variant in the seed handed to `/roadmap:create` (so the roadmap can
+reference them under "Context"). Step 4-bis council poll is identical
+to `vision.md`; pass the doc inventory + glossary as additional context
+in the council prompt.
 
 ## Steps
 
@@ -249,7 +262,7 @@ When the user states "X is implemented as Y" AND the claim is
 constraint), agent runs `view` / `grep` / `codebase-retrieval` to
 verify. Mismatch → single question block:
 
-> Your seed says X works via Y. The code (`path/to/file.php:123`) shows
+> Your seed says X works via Y. The code (`path/to/file:123`) shows
 > it via Z. Which is right?
 > 1. **Code is right, my seed was wrong** — recommended.
 > 2. Seed is right, code is stale — separate fix needed.

package/dist/agent-src/commands/commit/in-chunks.md

@@ -58,7 +58,7 @@ step 3 grouping rules:
 - Same feature / fix → one commit.
 - Migration + corresponding model/seeder → one commit.
 - Test + the class under test → one commit.
-- Style-only changes (ECS/Rector/formatter) → separate `style:` /
+- Style-only changes (formatter / auto-refactor output — ECS, Prettier, Ruff) → separate `style:` /
   `chore:` commit when mixed with logic changes.
 - Truly unrelated change → its own commit.
 

package/dist/agent-src/commands/commit.md

@@ -87,7 +87,7 @@ Rules for splitting:
 - **Do NOT split arbitrarily** — only split when the changes are logically independent.
 - **Prefer fewer, coherent commits** over many tiny ones.
 - **Tests always go with the code they test** unless there are many test-only changes.
-- **Style-only changes** (ECS/Rector formatting) may get their own `style:` or `chore:` commit
+- **Style-only changes** (formatter / auto-refactor output — ECS, Prettier, Ruff) may get their own `style:` or `chore:` commit
   if they are large and mixed with logic changes.
 
 ### 5. Present the commit plan (verbosity-gated)
@@ -123,6 +123,7 @@ Show the proposed commits as a numbered list, including which files go into each
 ```
 Proposed commits:
 
+(Laravel-project example)
 1. feat(DEV-1234): add absence type filter to working time report
    → app/Services/WorkingTimeService.php
    → app/Http/Controllers/WorkingTimeController.php

package/dist/agent-src/commands/cost/report.md

@@ -123,7 +123,8 @@ Orchestration telemetry (YYYY-MM, N dispatches):
 **Quality × cost pairing rule (council verdict):** the savings figure never
 renders without the quality columns. `first_pass_success_rate` and
 `escalation_rate` aggregate over lines carrying the `first_pass_success` /
-`escalated` booleans (see `orchestration-telemetry.md`); with ≥ 20 such lines
+`escalated` booleans (definitions: `orchestration-telemetry.md`
+§ Operationalization); with ≥ 20 such lines
 in the window the real rates render, below 20 render `n/a (n=<count>)`
 alongside the savings. `src/scripts/orchestration_savings_report.ts`
 implements this pairing — prefer running it over hand-summarizing.

package/dist/agent-src/commands/feature/dev.md

@@ -85,20 +85,20 @@ Do NOT use for single-line fixes, trivial changes, or urgent hotfixes.
 ### Phase 5: Implementation
 
 1. **Wait for explicit approval** before writing code.
-2. Create a task list with all impl steps.
+2. Create a task list with all implementation steps.
 3. Follow chosen architecture from Phase 4.
 4. Follow codebase conventions strictly (read guidelines).
 5. Update task list as progress is made.
-6. Write tests alongside the impl.
+6. Write tests alongside the implementation.
 
 ### Phase 6: Quality Review
 
-1. Review the impl for:
+1. Review the implementation for:
    - **Simplicity & DRY** — is the code as simple as possible?
    - **Correctness** — are there logic errors or missing edge cases?
    - **Conventions** — does it follow project patterns?
 2. Run quality tools:
-   - PHPStan (or equivalent)
+   - Static analysis (PHPStan, tsc, mypy — whichever the project runs)
    - Tests
    - Code style
 3. Present findings with severity levels.