@mmerterden/multi-agent-pipeline 14.2.2 → 15.0.0

package/CHANGELOG.md

@@ -16,6 +16,76 @@ Internal file-layout changes that don't affect the slash-command surface are sti
 
 ## [Unreleased]
 
+## [15.0.0] - 2026-08-13
+
+### Changed (BREAKING)
+
+- **Claude Code stack skills are plugin-only** (ADR-0009). The installer no longer
+  copies `shared/core` + `shared/external` into `~/.claude/skills`; the
+  `multi-agent-plugins` marketplace is Claude Code's only stack-skill source,
+  namespaced (`ai-ios-toolkit:<skill>`). Exactly two pipeline-owned compliance
+  catalogs stay local (`PIPELINE_LOCAL_SKILLS`). Old installs are migrated
+  manifest-scoped (`.external-skills-manifest.json`); user-authored dirs survive.
+- **Marketplace plugin names lose the `engineering` infix**:
+  `ai-<stack>-engineering-toolkit` -> `ai-<stack>-toolkit`. `enabledPlugins`
+  keys must migrate; `/multi-agent:stack` deletes the retired keys.
+- **`/multi-agent:ship` -> `/multi-agent:resume-local`** (the tail command opens a
+  PR but merges nothing, so "ship" oversold it). Prefs `global.ship` ->
+  `global.resumeLocal` via schema 2.6.0 (step migration + monolith, autoFix value
+  carried across the finish -> ship -> resume-local chain).
+
+### Added
+
+- `/multi-agent:stack` **multi-select**: several stacks per call (`ios backend`),
+  a native multi-select picker with no args, `web` alias, always-on
+  `ai-common-toolkit`, legacy-key cleanup, and a Copilot/Codex refresh offer.
+- Every marketplace plugin ships **`ai-<stack>-toolkit:help`**: a language-aware
+  catalog rendered live from `plugin.json` (drift-proof by construction).
+- `pipeline/scripts/_stack-routing.mjs`: the single routing table shared by
+  `build-stack-plugins.mjs`, the installers and the skills index.
+- Skills index entries carry `plugin` + `invokeAs`; `match-skills.mjs` returns
+  plugin-namespaced names. `skill-conformance.mjs` probes the marketplace
+  checkout and version-named plugin cache as skills roots.
+
+### Changed
+
+- Copilot CLI and Codex CLI local copies are **filtered to the enabled stacks**
+  (`pluginsToDeliver` + `partitionExternalSkillsByPlugins`); Codex previously
+  received the full catalog unconditionally and now also writes a delivery
+  manifest. One `/multi-agent:stack` selection governs all three hosts.
+- Prompt-context skill tables (phase 1/2/4), `humanizer` invocations,
+  `build-optimize` dispatches and the store-review references are
+  plugin-namespaced.
+- `DESC_CEILING` 82000 -> 82600 and phase-doc total 52750 -> 53100: the
+  namespaced tables and the new command descriptions did not fit ceilings that
+  were already at (or 193 bytes past) their limit; the always-on surface itself
+  shrank with the local copy gone.
+
+
+### Added
+
+- **`/multi-agent:complaint-analysis`** — customer-complaint / customer-reported-error
+  triage as a new one-shot ops command (51st command). Ingests complaints from
+  free-text paste, csv/xlsx/txt/json exports (`pipeline/lib/parse-complaints.sh`,
+  with built-in PII redaction: email / phone / card / national-id / PNR shapes),
+  Jira issues, or Confluence URLs; fetches Graylog evidence per
+  trxId/conversationId via the existing `fetch-graylog.sh`; correlates read-only
+  against the selected client + BFF repos (user-confirmed layer tagging:
+  ios / android / web / mobile-bff / web-bff); and classifies each complaint as
+  `client:<layer>` / `bff:<layer>` (root cause + citations + a fix plan grounded
+  in the existing architecture + a ready-to-run dev prompt), `core` (routing
+  recommendation to the backend core team, never a fix analysis), or
+  `insufficient-evidence`. Report dispatches to Local (default) / Confluence /
+  Jira behind a deterministic gate (`scripts/validate-complaint-doc.mjs`:
+  sections, verdict tokens, per-core routing entries, humanizer punctuation,
+  redaction-leak scan; smoke: `smoke-validate-complaint-doc.sh`). State contract:
+  `schemas/complaint-analysis-spec.schema.json`. Graylog-as-primary-evidence is a
+  documented, command-scoped exception to the advisory-only rule in
+  `features/external-context-injection.md`. `DESC_CEILING` raised 81000 -> 82000
+  (gate + pinned test together): the two new ~430-byte descriptions did not fit
+  in the ~25 bytes of headroom the surface had left; the average stays at 319
+  against the 420 ceiling.
+
 ## [14.2.2] - 2026-08-04
 
 ### Added
@@ -1911,7 +1981,7 @@ Single-source skill management, an upgraded `refactor` command, and two new Phas
 - **Component/Figma work is plugin-only.** The pipeline no longer bundles the
   `figma-ios` / `figma-android` / `figma-common` skill trees. Phase 3 dispatches
   component and Figma-to-code work to the per-stack marketplace plugins
-  (`ai-ios-engineering-toolkit` / `ai-android-engineering-toolkit`) via the Skill
+  (`ai-ios-toolkit` / `ai-android-toolkit`) via the Skill
   tool, so component skills live in one place. The 3-tier Figma design-access
   fallback still governs the analysis phase.
 - **Command layout migrated to `<name>/SKILL.md`.** Each subcommand is now its own
@@ -1952,7 +2022,7 @@ Skill mining + an install-safety fix.
   write-if-missing: new baseline rules are added, existing local rules are
   preserved untouched. New `copyDir({ skipExisting })` option backs this.
 - **Two skills mined from the corporate iOS toolkit and genericized into
-  `ai-common-engineering-toolkit` (v0.1.2):**
+  `ai-common-toolkit` (v0.1.2):**
   - `skill-creator` — the house rules for authoring skills (description-first
     discovery, lean SKILL.md as an index, progressive disclosure, reference vs
     workflow vs tool layers, no-prefix naming, grow-from-failure), with a
@@ -1987,7 +2057,7 @@ operating systems, all additive and opt-in.
   no-progress stall, an identical repeated failure, a rework storm, cost drift
   past the `costBudget` ceiling, or a merge/rebase conflict. Continuing
   unattended off the happy path is the less safe choice.
-- **Three technique skills** added to `ai-common-engineering-toolkit`
+- **Three technique skills** added to `ai-common-toolkit`
   (v0.1.1): `council` (multi-voice adversarial decision), `search-first`
   (research-before-coding with an adopt/extend/compose/build matrix), and
   `agent-introspection-debugging` (capture -> diagnose -> contained-recovery ->
@@ -2343,7 +2413,7 @@ Fable 5 restored as the top model tier; `stack-swap` fully removed; setup gains
   setup detects the project stack from markers (`.xcodeproj`/`Package.swift` → iOS,
   `build.gradle` → Android, `package.json`+react → Frontend, `requirements.txt`/
   `pyproject.toml` → Backend) and enables the matching marketplace plugin plus the
-  always-on `ai-common-engineering-toolkit`. **No clear marker → default iOS.** So a
+  always-on `ai-common-toolkit`. **No clear marker → default iOS.** So a
   fresh install works out of the box and a repo at any org gets its correct stack.
 
 ### Changed
@@ -2377,7 +2447,7 @@ Stack skills move from the local `stack-swap` mechanic to versioned marketplace
   `plugin.json` `skills[]`, and **bumps the patch version of any plugin whose
   skill set changed**. Idempotent (`--dry-run` supported); a no-op run bumps
   nothing. Cross-stack skills (accessibility audit, humanizer, Firebase) route
-  to `ai-common-engineering-toolkit`; Apple/Xcode-only skills stay in the iOS
+  to `ai-common-toolkit`; Apple/Xcode-only skills stay in the iOS
   plugin. This is the version-based-management backbone: the pipeline is the
   single authoring source, the marketplace is a derived, versioned artifact.
 - **`multi-agent:sync` Step 3c (PLUGINS)** — runs the generator, then commits +
@@ -2387,7 +2457,7 @@ Stack skills move from the local `stack-swap` mechanic to versioned marketplace
 
 - **`/multi-agent:stack`** — selecting a stack no longer moves skill directories
   via `stack-swap.sh`. It now enables the matching marketplace plugin(s) (stack
-  toolkit + `ai-common-engineering-toolkit`) in the current repo's
+  toolkit + `ai-common-toolkit`) in the current repo's
   `.claude/settings.json` `enabledPlugins`, disabling the stack toolkits that
   don't apply. Declarative, per-repo, versioned. No SessionStart hook.
 - **`/multi-agent:update`** — added a step that refreshes the plugin marketplace

package/README.md

@@ -17,13 +17,18 @@ Runs natively on Claude Code, Copilot CLI and Codex CLI. macOS / Linux / Windows
 ## Quick Start
 
 ```bash
-# from the public registry (no auth)
-npx @mmerterden/multi-agent-pipeline install --all   # Claude Code + Copilot CLI + Codex CLI
+# from the public registry (no auth) - pick the CLIs you actually use
+npx @mmerterden/multi-agent-pipeline install --claude    # Claude Code only (default)
+npx @mmerterden/multi-agent-pipeline install --copilot   # Copilot CLI only
+npx @mmerterden/multi-agent-pipeline install --codex     # Codex CLI only
+npx @mmerterden/multi-agent-pipeline install --all       # all three
 
 # then, once:
 /multi-agent:setup          # keychain token scan + git identity + default stack
 ```
 
+Tool flags combine (`--claude --codex`). With no tool flag at all, the installer targets Claude Code only. Other flags: `--dry-run` (show what would be written, write nothing), `--platform=ios|android|all` (skip the stack skills you do not need), `--link` (symlink instead of copy, for local development).
+
 Run a task - the input type is auto-detected:
 
 ```bash
@@ -41,6 +46,8 @@ Add `autopilot` to skip confirmations, `--dev` for the fast dev-only path, or `-
 
 Update later with `/multi-agent:update`. Uninstall (tokens preserved) with `npx @mmerterden/multi-agent-pipeline uninstall`.
 
+**Stack skills are marketplace plugins.** On Claude Code the `ai-<stack>-toolkit` plugins (`multi-agent-plugins` marketplace) are the only stack-skill source - nothing is copied into `~/.claude/skills`. Pick the active stack(s) per repo with `/multi-agent:stack` (multi-select: `ios backend`, or a native picker with no args); each plugin ships a language-aware catalog at `ai-<stack>-toolkit:help`. Copilot CLI and Codex CLI have no plugin loader, so they receive a local copy filtered to the same enabled stacks.
+
 ## How it works
 
 One command runs 8 phases, with a gate between the risky ones:
@@ -66,11 +73,11 @@ The discipline behind all of this - bounded loops, evidence gates, token-budgete
 | Autopilot | `/multi-agent:autopilot "task"` | All 8 phases, no confirmations |
 | Dev | `/multi-agent:dev "task"` | Init → Dev → Review → Test → Commit → Report |
 | Local | `/multi-agent:local "task"` | Full pipeline, current branch (no worktree) |
-| Ship | `/multi-agent:ship` | Run the review→test→commit→report tail over local work |
+| Ship | `/multi-agent:resume-local` | Run the review→test→commit→report tail over local work |
 | Audit | `/multi-agent:design-check` | Mock-mode vs Figma conformance, local-only |
 | Audit | `/multi-agent:testflight-validation` | Pre-submission gates for a TestFlight build: static archive audit → Apple's `altool --validate-app` → Review-Guidelines check. Validates only, never uploads |
 
-Helpers: `setup`, `status`, `resume #N`, `kill #N`, `garbage-collect`, `prune-logs`, `purge`, `review`, `test`, `channels`, `stack`, `update`, `sync`, `refactor`, `jira`, `issue`, `analysis`, `create-jira`, `save`, `routines`, `forget`. 49 commands in all  -  full list: `/multi-agent:help`.
+Helpers: `setup`, `status`, `resume #N`, `kill #N`, `garbage-collect`, `prune-logs`, `purge`, `review`, `test`, `channels`, `stack`, `update`, `sync`, `refactor`, `jira`, `issue`, `analysis`, `create-jira`, `save`, `routines`, `forget`. 50 commands in all  -  full list: `/multi-agent:help`.
 
 ## Stacks
 
@@ -84,17 +91,17 @@ This enables the matching plugin (+ the shared `ai-common` plugin) in the repo's
 
 ## Tool support
 
-The pipeline runs natively on **Claude Code**, **Copilot CLI** and **Codex CLI** - all three install from the same `pipeline/` source and get the same 49 commands.
+The pipeline runs natively on **Claude Code**, **Copilot CLI** and **Codex CLI** - all three install from the same `pipeline/` source and get the same 50 commands.
 
 | Tool | Flag | What it installs |
 |---|---|---|
 | Claude Code | `--claude` (default) | slash commands + skills + agents + `PreToolUse` secret-scan hook |
-| Copilot CLI | `--copilot` | instructions + 49 sub-command skills + scripts |
-| Codex CLI | `--codex` | one router skill + 49 specs as refs + 8 agent TOML + `AGENTS.md` block + `codex mcp add` |
+| Copilot CLI | `--copilot` | instructions + 50 sub-command skills + scripts |
+| Codex CLI | `--codex` | one router skill + 50 specs as refs + 8 agent TOML + `AGENTS.md` block + `codex mcp add` |
 
 Filter skills by stack with `--platform=ios\|android\|all`.
 
-**Why Codex gets one skill and not 49.** Codex assembles every discovered skill's name
+**Why Codex gets one skill and not 50.** Codex assembles every discovered skill's name
 and description into a single prompt block and drops entries when it overflows, with no
 error. Measured on 0.145: installing one plugin that declares 142 skills surfaced only
 75 of them and evicted an unrelated user skill. So on Codex the pipeline ships a single

package/README.tr.md

@@ -17,13 +17,18 @@ Claude Code, Copilot CLI ve Codex CLI üzerinde native çalışır. macOS / Linu
 ## Hızlı Başlangıç
 
 ```bash
-# public registry'den (auth gerekmez)
-npx @mmerterden/multi-agent-pipeline install --all   # Claude Code + Copilot CLI + Codex CLI
+# public registry'den (auth gerekmez) - sadece kullandığın CLI'ları seç
+npx @mmerterden/multi-agent-pipeline install --claude    # sadece Claude Code (varsayılan)
+npx @mmerterden/multi-agent-pipeline install --copilot   # sadece Copilot CLI
+npx @mmerterden/multi-agent-pipeline install --codex     # sadece Codex CLI
+npx @mmerterden/multi-agent-pipeline install --all       # üçü birden
 
 # sonra, bir kere:
 /multi-agent:setup          # keychain token taraması + git kimliği + varsayılan stack
 ```
 
+Tool flag'leri birleştirilebilir (`--claude --codex`). Hiç tool flag'i verilmezse installer sadece Claude Code'u hedefler. Diğer flag'ler: `--dry-run` (ne yazılacağını gösterir, hiçbir şey yazmaz), `--platform=ios|android|all` (ihtiyacın olmayan stack skill'lerini atlar), `--link` (kopyalamak yerine symlink, lokal geliştirme için).
+
 Bir görev çalıştır - girdi tipi otomatik algılanır:
 
 ```bash
@@ -41,6 +46,8 @@ Onayları atlamak için `autopilot`, hızlı dev-only yol için `--dev`, ya da w
 
 Sonra `/multi-agent:update` ile güncelle. Kaldırmak için (tokenlar korunur) `npx @mmerterden/multi-agent-pipeline uninstall`.
 
+**Stack skill'leri marketplace plugin'leridir.** Claude Code'da `ai-<stack>-toolkit` plugin'leri (`multi-agent-plugins` marketplace) tek stack-skill kaynağıdır - `~/.claude/skills` altına kopya konmaz. Repo başına aktif stack'leri `/multi-agent:stack` ile seç (çoklu seçim: `ios backend`, ya da argümansız native picker); her plugin `ai-<stack>-toolkit:help` ile dilinde katalog sunar. Copilot CLI ve Codex CLI'da plugin loader olmadığından onlara aynı enabled stack'lere filtreli lokal kopya iner.
+
 ## Nasıl çalışır
 
 Tek komut 8 fazı çalıştırır, riskli olanlar arasında bir kapı ile:
@@ -66,11 +73,11 @@ Bunun arkasındaki disiplin - sınırlı loop'lar, kanıt kapıları, token-büt
 | Autopilot | `/multi-agent:autopilot "task"` | Tüm 8 faz, onaysız |
 | Dev | `/multi-agent:dev "task"` | Init → Dev → Review → Test → Commit → Report |
 | Local | `/multi-agent:local "task"` | Tam pipeline, mevcut branch (worktree yok) |
-| Ship | `/multi-agent:ship` | Lokal iş üzerinde review→test→commit→report kuyruğunu çalıştır |
+| Ship | `/multi-agent:resume-local` | Lokal iş üzerinde review→test→commit→report kuyruğunu çalıştır |
 | Audit | `/multi-agent:design-check` | Mock-mode vs Figma uygunluğu, yalnızca lokal |
 | Audit | `/multi-agent:testflight-validation` | TestFlight build için pre-submission kapıları: statik archive denetimi → Apple'ın `altool --validate-app`'i → Review-Guidelines kontrolü. Yalnızca doğrular, asla yüklemez |
 
-Yardımcılar: `setup`, `status`, `resume #N`, `kill #N`, `garbage-collect`, `prune-logs`, `purge`, `review`, `test`, `channels`, `stack`, `update`, `sync`, `refactor`, `jira`, `issue`, `analysis`, `create-jira`, `save`, `routines`, `forget`. Toplam 49 komut - tam liste: `/multi-agent:help`.
+Yardımcılar: `setup`, `status`, `resume #N`, `kill #N`, `garbage-collect`, `prune-logs`, `purge`, `review`, `test`, `channels`, `stack`, `update`, `sync`, `refactor`, `jira`, `issue`, `analysis`, `create-jira`, `save`, `routines`, `forget`. Toplam 50 komut - tam liste: `/multi-agent:help`.
 
 ## Stack'ler
 
@@ -84,17 +91,17 @@ Bu, ilgili plugin'i (+ ortak `ai-common` plugin'ini) repo'nun `.claude/settings.
 
 ## Araç desteği
 
-Pipeline **Claude Code**, **Copilot CLI** ve **Codex CLI** üzerinde native çalışır - üçü de aynı `pipeline/` kaynağından kurulur ve aynı 49 komutu alır.
+Pipeline **Claude Code**, **Copilot CLI** ve **Codex CLI** üzerinde native çalışır - üçü de aynı `pipeline/` kaynağından kurulur ve aynı 50 komutu alır.
 
 | Araç | Bayrak | Ne kurar |
 |---|---|---|
 | Claude Code | `--claude` (varsayılan) | slash komutları + skill'ler + agent'lar + `PreToolUse` secret-scan hook'u |
-| Copilot CLI | `--copilot` | talimatlar + 49 alt-komut skill'i + script'ler |
-| Codex CLI | `--codex` | bir router skill + ref olarak 49 spec + 8 agent TOML + `AGENTS.md` bloğu + `codex mcp add` |
+| Copilot CLI | `--copilot` | talimatlar + 50 alt-komut skill'i + script'ler |
+| Codex CLI | `--codex` | bir router skill + ref olarak 50 spec + 8 agent TOML + `AGENTS.md` bloğu + `codex mcp add` |
 
 Skill'leri stack'e göre filtrele: `--platform=ios\|android\|all`.
 
-**Codex neden 49 değil de tek bir skill alıyor.** Codex, keşfettiği her skill'in adını
+**Codex neden 50 değil de tek bir skill alıyor.** Codex, keşfettiği her skill'in adını
 ve açıklamasını tek bir prompt bloğuna toplar ve blok taştığında girdileri hatasızca
 düşürür. 0.145 üzerinde ölçüldü: 142 skill deklare eden bir plugin kurulduğunda sadece
 75'i yüzeye çıktı ve alakasız bir kullanıcı skill'i tahliye edildi. Bu yüzden Codex'te

package/docs/FIGMA_PIPELINE.md

@@ -2,7 +2,7 @@
 
 End-to-end guide for generating production-ready UI components from Figma designs through the multi-agent pipeline. Covers iOS (SwiftUI) and Android (Jetpack Compose) with the same workflow shape.
 
-> Note: Component and Figma-to-code skills are no longer bundled in this repo. Phase 3 dispatches component work to the per-stack marketplace plugins (`ai-ios-engineering-toolkit` / `ai-android-engineering-toolkit` in the `multi-agent-plugins` marketplace) via the Skill tool. The workflow, subphases, and skill names described below are what those plugin skills run internally, not pipeline-bundled code. The 3-tier Figma design-access fallback still governs the analysis phase in this repo.
+> Note: Component and Figma-to-code skills are no longer bundled in this repo. Phase 3 dispatches component work to the per-stack marketplace plugins (`ai-ios-toolkit` / `ai-android-toolkit` in the `multi-agent-plugins` marketplace) via the Skill tool. The workflow, subphases, and skill names described below are what those plugin skills run internally, not pipeline-bundled code. The 3-tier Figma design-access fallback still governs the analysis phase in this repo.
 
 ## What you get
 
@@ -15,7 +15,7 @@ End-to-end guide for generating production-ready UI components from Figma design
 
 ## How it runs
 
-When multi-agent Phase 0 classifies a task as `component` (Figma URL in description, or figma-driven instruction path), Phase 3 delegates the entire phase to the enabled `ai-<platform>-engineering-toolkit` marketplace plugin's component skill (`create-component`, fallback `create-ui-component`) via the Skill tool. Component skills live in the plugin marketplace, not the pipeline. The dispatch layer records a coarse component-build row in `state.phases["3"].subphases[]` - multi-agent's phase-tracker reads that array with no special case. The subphase list below describes the flow the plugin skill runs internally.
+When multi-agent Phase 0 classifies a task as `component` (Figma URL in description, or figma-driven instruction path), Phase 3 delegates the entire phase to the enabled `ai-<platform>-toolkit` marketplace plugin's component skill (`create-component`, fallback `create-ui-component`) via the Skill tool. Component skills live in the plugin marketplace, not the pipeline. The dispatch layer records a coarse component-build row in `state.phases["3"].subphases[]` - multi-agent's phase-tracker reads that array with no special case. The subphase list below describes the flow the plugin skill runs internally.
 
 ### Internal phase order
 
@@ -122,7 +122,7 @@ Global settings that affect the figma pipeline:
 
 | Symptom | Likely cause | Fix |
 |---|---|---|
-| Phase 3 halts with "plugin not enabled" | The `ai-<platform>-engineering-toolkit` plugin is not enabled in this repo | Enable it in the repo's `.claude/settings.local.json` (`"ai-ios-engineering-toolkit@<marketplace>": true`) and reload the session |
+| Phase 3 halts with "plugin not enabled" | The `ai-<platform>-toolkit` plugin is not enabled in this repo | Enable it in the repo's `.claude/settings.local.json` (`"ai-ios-toolkit@<marketplace>": true`) and reload the session |
 | `multi-agent:create-component` not found | Marketplace not installed / plugin disabled | Install the marketplace and enable the platform toolkit; dispatch tries `create-component` then `create-ui-component` |
 | Wiki adapter failure | Remote unreachable (separate-repo mode) | Adapter caches pending output; next run retries. Non-blocking - Phase 7 continues. |
 | Jira auto-create hits 5xx | Transient Jira outage | Non-blocking - pipeline continues without link. User can run `/multi-agent:channels` post-hoc to add it. |

package/docs/adr/0006-skills-core-external-split.md

@@ -1,6 +1,6 @@
 # 6. `shared/core/` vs `shared/external/` source organization
 
-**Status:** Accepted · 2026-04 (v5.3.3)
+**Status:** Accepted · 2026-04 (v5.3.3) · The "install destination is unchanged" clause is superseded by [ADR-0009](./0009-claude-stack-skills-plugin-only.md) (v15.0.0): Claude Code no longer receives the external tree; Copilot/Codex receive a stack-filtered subset.
 
 ## Context
 

package/docs/adr/0009-claude-stack-skills-plugin-only.md

@@ -0,0 +1,31 @@
+# 9. Claude Code stack skills are plugin-only; the local copy is retired
+
+**Status:** Accepted · 2026-08 (v15.0.0) · Supersedes the "install destination is unchanged" clause of ADR-0006
+
+## Context
+
+ADR-0006 split the source tree into `shared/core/` + `shared/external/` but kept the install unchanged: both trees flattened into `~/.claude/skills/`. Since then the same external catalog also became the authoring source for the `multi-agent-plugins` marketplace (`build-stack-plugins.mjs` derives each plugin's `knowledge/` layer from it), and Claude Code loads those plugins natively, namespaced (`ai-ios-toolkit:swiftui-pro`).
+
+That left Claude Code receiving every stack skill **twice**:
+
+- 110 of 157 local skill dirs duplicated the enabled plugins byte-for-byte - ~10k tokens of duplicate descriptions per session, 4.4 MB on disk.
+- The local copy shadowed the plugin the moment it went stale, and it *was* stale in practice: the plugin cache and the local copy advanced on different schedules.
+- The duplication hid real bugs: `skill-conformance.mjs` and `match-skills.mjs` only knew the local root, so they kept working by accident - against the stale copy.
+
+## Decision
+
+1. **Claude Code receives no local copy of the stack skills.** The marketplace plugins (`ai-<stack>-toolkit`, enablement managed by `/multi-agent:stack`, multi-select) are its only stack-skill source. Prompt tables, `Skill` invocations and the skills index all use the plugin-namespaced name.
+2. **Exactly two skills stay local on Claude Code**: `apple-archive-compliance` and `google-play-compliance` (`PIPELINE_LOCAL_SKILLS` in `install/claude.mjs`). They are pipeline-owned compliance catalogs, kept out of the public marketplace by policy (they may carry audit rules not meant for it), and `uninstall.mjs` already treats exactly this pair as pipeline-owned.
+3. **`shared/external/` stays in the repo as the authoring source.** It feeds `build-stack-plugins.mjs` and the Copilot/Codex copies; it is simply never copied to `~/.claude/skills` again. Installs migrate old machines by removing the dirs named in `.external-skills-manifest.json` (manifest-scoped - user-authored dirs survive).
+4. **Copilot CLI and Codex CLI keep local copies** (no plugin loader / silently truncating skill list), now **filtered to the enabled stacks**: `pluginsToDeliver` reads `enabledPlugins` from `~/.claude/settings.json` (fallback `--platform`), `partitionExternalSkillsByPlugins` routes each skill through the same `_stack-routing.mjs` table the marketplace is built with. One stack selection governs all three hosts.
+5. **Scripts that resolve skills know the plugin roots**: `skill-conformance.mjs` probes the marketplace checkout and version-named plugin cache (newest first) alongside the install roots; `build-skills-index.mjs` stamps `plugin` + `invokeAs` per entry so `match-skills.mjs` returns namespaced names.
+
+## Consequences
+
+Positive: one source of truth per skill at runtime; ~13.5k tokens/session and ~5.9 MB reclaimed on Claude Code; a disabled stack is genuinely absent instead of half-present; the smoke suite now asserts the compliance-only contract (`smoke-install-layout.sh` 5b2b) and the filter's both directions (5b2).
+
+Negative / accepted:
+
+- A stack not enabled in `enabledPlugins` is invisible to Claude Code until `/multi-agent:stack` enables it. This is the intended behavior, not a regression.
+- `dynamicSkillLoading`'s index now names skills the local tree does not hold; consumers must use `invokeAs`, not a filesystem path.
+- Copilot/Codex copies go stale when the stack selection changes until `install --copilot --codex` is re-run; `/multi-agent:stack` offers that refresh after every change.

package/docs/adr/README.md

@@ -18,6 +18,7 @@ Format: lightly adapted from [Michael Nygard's ADR template](https://cognitect.c
 | [0006](./0006-skills-core-external-split.md) | `shared/core/` vs `shared/external/` source org    | Accepted |
 | [0007](./0007-multi-tool-adapter-framework.md) | Multi-tool adapter framework + token-preserving uninstall | Superseded by v10.7.0 (adapters removed; Claude Code + Copilot CLI only) |
 | [0008](./0008-installer-modularization-and-secret-leak-defense.md) | Installer modularization + secret-leak defense | Accepted (amended v10.7.0: adapter module removed) |
+| [0009](./0009-claude-stack-skills-plugin-only.md) | Claude Code stack skills are plugin-only; local copy retired | Accepted (supersedes ADR-0006 install-destination clause) |
 
 ## Writing a New ADR
 

package/docs/architecture.md

@@ -71,7 +71,7 @@ graph TD
 
 ## Figma SubPhase Integration (Phase 3)
 
-When a task is classified `component`, Phase 3 dispatches to the marketplace component plugin (`ai-<platform>-engineering-toolkit`) via the Skill tool. Component skills are not bundled in this repo; the subphases below describe the flow the plugin skill runs internally:
+When a task is classified `component`, Phase 3 dispatches to the marketplace component plugin (`ai-<platform>-toolkit`) via the Skill tool. Component skills are not bundled in this repo; the subphases below describe the flow the plugin skill runs internally:
 
 ```mermaid
 graph TD
@@ -117,7 +117,7 @@ graph TB
     end
 
     subgraph "Pipeline Specs"
-        CMD[commands/<br/>49 command files]
+        CMD[commands/<br/>50 command files]
         AGT[agents/<br/>8 agent personas]
         RUL[rules/<br/>12 domain rules]
         PHS[multi-agent-refs/phases/<br/>phase specs + contracts]
@@ -163,14 +163,14 @@ User Input → Phase 0 (Init)
 ## Ecosystem Sync Targets
 
 `/multi-agent:sync` actually walks **five** targets, not the three shown in earlier
-revisions of this diagram — Codex CLI and the two independently-shipped repos
+revisions of this diagram - Codex CLI and the two independently-shipped repos
 (`multi-agent-plugins`, `dev-toolkit-mcp`) are real sync targets too:
 
 ```mermaid
 graph TD
     CC["Claude Code<br/>(source of truth)"]
-    COP["Copilot CLI<br/>(instructions + 49 skills)"]
-    COD["Codex CLI<br/>(1 router skill + 49 refs)"]
+    COP["Copilot CLI<br/>(instructions + 50 skills)"]
+    COD["Codex CLI<br/>(1 router skill + 50 refs)"]
     REPO["Pipeline Repo<br/>(npm package)"]
     WEB["Website"]
     PLUGREPO["multi-agent-plugins<br/>(5 stack plugins, own repo)"]
@@ -189,6 +189,6 @@ graph TD
     DTK -->|npm publish| NPM2["GitHub Packages (private)"]
 ```
 
-Full detail on how these three repos compose at install time and at run time —
+Full detail on how these three repos compose at install time and at run time -
 including the Phase 3 → plugin dispatch contract and the Phase 5 → dev-toolkit MCP
-contract — lives in [`docs/ecosystem.md`](./ecosystem.md).
+contract - lives in [`docs/ecosystem.md`](./ecosystem.md).

package/docs/ecosystem.md

@@ -5,7 +5,7 @@ separately, wired together at install time and at run time:
 
 | Repo | What it owns | Ships as |
 |---|---|---|
-| **`multi-agent-pipeline`** (this repo) | Orchestration: the 8-phase flow, the 49 slash commands, quality gates, review/triage, cross-CLI parity | npm package (`@mmerterden/multi-agent-pipeline`), installs itself onto Claude Code / Copilot CLI / Codex CLI |
+| **`multi-agent-pipeline`** (this repo) | Orchestration: the 8-phase flow, the 50 slash commands, quality gates, review/triage, cross-CLI parity | npm package (`@mmerterden/multi-agent-pipeline`), installs itself onto Claude Code / Copilot CLI / Codex CLI |
 | **`multi-agent-plugins`** | Stack knowledge: per-platform component/lifecycle skills (iOS, Android, Frontend, Backend) + shared knowledge | Claude Code marketplace, 5 independently-versioned plugins |
 | **`dev-toolkit-mcp`** | The pipeline's hands on devices and browsers: 80 MCP tools across 6 categories (simulator/emulator control, accessibility audit, store compliance, web automation, Figma-vs-mock design audit, an agent-DSL batch runner) | npm package, registered as a standard stdio MCP server on every host |
 
@@ -18,17 +18,17 @@ Either can be swapped or removed without touching the other two's source.
 graph LR
     subgraph PIPE ["multi-agent-pipeline (orchestrator)"]
         direction TB
-        PHASES["8 phases · 49 commands"]
+        PHASES["8 phases · 50 commands"]
         GATES["deterministic gates + review triage"]
     end
 
     subgraph PLUG ["multi-agent-plugins (stack knowledge)"]
         direction TB
-        IOSP["ai-ios-engineering-toolkit"]
-        ANDP["ai-android-engineering-toolkit"]
-        FEP["ai-frontend-engineering-toolkit"]
+        IOSP["ai-ios-toolkit"]
+        ANDP["ai-android-toolkit"]
+        FEP["ai-frontend-toolkit"]
         BEP["ai-backend-toolkit"]
-        COMP["ai-common-engineering-toolkit"]
+        COMP["ai-common-toolkit"]
     end
 
     subgraph DTK ["dev-toolkit-mcp (device/browser hands)"]
@@ -63,8 +63,8 @@ only those:
 graph TD
     CC["Claude Code<br/>~/.claude/commands/multi-agent/<br/>(source of truth)"]
 
-    CC -->|"Step 2: copy + reformat<br/>49 sub-command skills"| COP["Copilot CLI<br/>~/.copilot/skills/"]
-    CC -->|"Step 2b: transform<br/>(install.js --codex)"| COD["Codex CLI<br/>1 router skill + 49 refs<br/>+ 8 agent TOML"]
+    CC -->|"Step 2: copy + reformat<br/>50 sub-command skills"| COP["Copilot CLI<br/>~/.copilot/skills/"]
+    CC -->|"Step 2b: transform<br/>(install.js --codex)"| COD["Codex CLI<br/>1 router skill + 50 refs<br/>+ 8 agent TOML"]
     CC -->|"Step 3: genericize<br/>(strip personal data)"| REPO["multi-agent-pipeline repo<br/>pipeline/"]
     CC -->|"Step 4: version + feature sync"| WEB["Website<br/>projects.ts / i18n.tsx"]
 
@@ -85,18 +85,18 @@ graph TD
 ```
 
 **Why `dev-toolkit-mcp` is drawn differently.** The other four targets are *derived*
-from the Claude Code source — sync writes their content. `dev-toolkit-mcp` is not:
+from the Claude Code source - sync writes their content. `dev-toolkit-mcp` is not:
 it's a separate codebase developed on its own schedule. Sync's Step 3d only
 *detects* whether it moved (dirty tree, unpushed commits, untagged version), runs
-**its own** gate suite, and ships it — commit, tag, `npm publish`. If the pipeline
+**its own** gate suite, and ships it - commit, tag, `npm publish`. If the pipeline
 needs a tool that toolkit doesn't have yet, that's a two-repo change: add the tool
 in `dev-toolkit-mcp`, ship it, then bump the minimum version pin back in
-`cross-cli-contract.md` (see §4).
+`cross-cli-contract.md` (see section 4).
 
 **Also not generated: the plugins' own authored skills.** `build-stack-plugins.mjs`
 only rebuilds each plugin's `knowledge/` folder from `pipeline/skills/shared/external/`.
-The plugins' lifecycle skills — `create-component`, `evolve-component`,
-`figma-utility`, `code-connect`, `branch-and-pr`, `fix-bug`, and the rest — are
+The plugins' lifecycle skills - `create-component`, `evolve-component`,
+`figma-utility`, `code-connect`, `branch-and-pr`, `fix-bug`, and the rest - are
 hand-authored *inside* `multi-agent-plugins` and are never touched by sync.
 
 ---
@@ -105,13 +105,13 @@ hand-authored *inside* `multi-agent-plugins` and are never touched by sync.
 
 ```mermaid
 graph TD
-    EXT["pipeline/skills/shared/external/<br/>150 skills — single authoring source<br/>(the pipeline's own phases read these too)"]
+    EXT["pipeline/skills/shared/external/<br/>150 skills - single authoring source<br/>(the pipeline's own phases read these too)"]
 
-    EXT -->|"cross-stack skills"| COMMONP["ai-common-engineering-toolkit<br/>10 skills · v0.2.3"]
-    EXT -->|"Apple/Xcode-only"| IOSP["ai-ios-engineering-toolkit<br/>145 skills · v0.6.0"]
-    EXT -->|"Android/Kotlin-only"| ANDP["ai-android-engineering-toolkit<br/>29 skills · v0.1.3"]
+    EXT -->|"cross-stack skills"| COMMONP["ai-common-toolkit<br/>10 skills · v0.2.3"]
+    EXT -->|"Apple/Xcode-only"| IOSP["ai-ios-toolkit<br/>145 skills · v0.6.0"]
+    EXT -->|"Android/Kotlin-only"| ANDP["ai-android-toolkit<br/>29 skills · v0.1.3"]
     EXT -->|"backend-only"| BEP["ai-backend-toolkit<br/>32 skills · v0.1.4"]
-    EXT -->|"web/frontend-only"| FEP["ai-frontend-engineering-toolkit<br/>24 skills · v0.1.3"]
+    EXT -->|"web/frontend-only"| FEP["ai-frontend-toolkit<br/>24 skills · v0.1.3"]
 
     COMMONP --> BUMP{"skill set<br/>changed?"}
     IOSP --> BUMP
@@ -130,7 +130,7 @@ A skill counted in more than one platform plugin (a cross-stack knowledge skill
 plus, say, an iOS-specific one) is why the plugins' skill counts sum to more than
 the 150-skill source: `ai-common` skills are vendored into every stack plugin's
 `knowledge/`, not deduplicated across them. Versioning is per-plugin and
-patch-only from this generator — a repo enabling only `ai-ios-engineering-toolkit`
+patch-only from this generator - a repo enabling only `ai-ios-toolkit`
 never pulls an Android-only change.
 
 **Consumption is pull, not push.** A consumer repo enables a stack plugin once
@@ -150,9 +150,9 @@ measurements behind this table):
 
 | | Claude Code | Copilot CLI | Codex CLI |
 |---|---|---|---|
-| **Pipeline commands** | 49 slash-command skills, native | 49 skills, `multi-agent-{cmd}` naming, copied in | 1 router skill (`multi-agent`) + 49 command specs as reference files — Codex silently truncates its skills block past a few dozen entries, so sub-commands are not peer skills here |
-| **Stack plugins** | Marketplace plugin, loaded natively, resolved by `.claude/settings.json` enabled-list | Enabled plugin's authored skills copied flat into `~/.copilot/skills/`; `knowledge/` **not** re-copied (already delivered via `shared/external`) | Copied as reference files under `~/.codex/multi-agent-refs/skills/`, plugin-prefixed on name clash (e.g. `architecture` → `ai-ios-engineering-toolkit-architecture`) |
-| **Component dispatch (Phase 3)** | Marketplace plugin's `create-component`/`create-screen` skill via the Skill tool | No plugin loader — falls back to local frozen `figma-*` skill copies | Not part of the enforced parity axis; classification + state-shape must match, skill *inventory* does not |
+| **Pipeline commands** | 50 slash-command skills, native | 50 skills, `multi-agent-{cmd}` naming, copied in | 1 router skill (`multi-agent`) + 50 command specs as reference files - Codex silently truncates its skills block past a few dozen entries, so sub-commands are not peer skills here |
+| **Stack plugins** | Marketplace plugin, loaded natively, resolved by `.claude/settings.json` enabled-list | Enabled plugin's authored skills copied flat into `~/.copilot/skills/`; `knowledge/` **not** re-copied (already delivered via `shared/external`) | Copied as reference files under `~/.codex/multi-agent-refs/skills/`, plugin-prefixed on name clash (e.g. `architecture` → `ai-ios-toolkit-architecture`) |
+| **Component dispatch (Phase 3)** | Marketplace plugin's `create-component`/`create-screen` skill via the Skill tool | No plugin loader - falls back to local frozen `figma-*` skill copies | Not part of the enforced parity axis; classification + state-shape must match, skill *inventory* does not |
 | **dev-toolkit-mcp** | `claude mcp add dev-toolkit -- npx -y @mmerterden/dev-toolkit-mcp` | `copilot mcp add dev-toolkit -- npx -y @mmerterden/dev-toolkit-mcp` | `codex mcp add dev-toolkit -- npx -y @mmerterden/dev-toolkit-mcp` (skipped with a warning if `codex` isn't on `PATH`) |
 
 `smoke-cross-cli-behavior.sh` and `smoke-codex-install.sh` gate the axes that **do**
@@ -170,7 +170,7 @@ other:
 graph TD
     START["Task running: Phase 3 (Dev)"]
     START -->|"taskType !== component"| TDD["Standard TDD loop<br/>(pipeline's own code)"]
-    START -->|"taskType === component<br/>+ figmaUrl present"| VALIDATE["ai-ios-engineering-toolkit:figma-validate<br/>(registry, Code Connect, token compliance)"]
+    START -->|"taskType === component<br/>+ figmaUrl present"| VALIDATE["ai-ios-toolkit:figma-validate<br/>(registry, Code Connect, token compliance)"]
     VALIDATE -->|pass| DISPATCH["Skill tool →<br/>create-component / create-screen<br/>/ evolve-component (dual-name fallback)"]
     VALIDATE -->|fail| HALT1["halt Phase 3, surface why"]
     DISPATCH --> REPORT1["plugin returns build/test status →<br/>dispatch layer writes state.phases['3'].subphases[]"]
@@ -196,9 +196,9 @@ hard stop, per `component-dispatch.md`).
 **Phase 5 (and design-check / store-ready) → dev-toolkit** is a long-lived MCP
 session, not a one-shot call: the same stdio server process answers many tool
 calls across a phase (boot simulator once, then screenshot/tap/screenshot/tap...).
-Several pipeline skills pin a **minimum toolkit version** for a specific tool —
+Several pipeline skills pin a **minimum toolkit version** for a specific tool -
 e.g. `apple-archive-compliance` requires `ios_app_store_audit` from
-`dev-toolkit-mcp ≥ v2.9.0` — enforced in `cross-cli-contract.md` and checked by
+`dev-toolkit-mcp ≥ v2.9.0` - enforced in `cross-cli-contract.md` and checked by
 `/multi-agent:sync` Step 3d before any dev-toolkit release ships (a version bump
 that drops or renames a tool a pipeline skill depends on is a **major** bump, by
 that step's own contract).
@@ -221,8 +221,8 @@ that step's own contract).
 - **Pipeline ↔ plugins boundary = Skill tool, one direction.** The pipeline
   classifies (`taskType`, `componentScope`) and tracks state; it never reads or
   writes plugin-internal files. This is why a corporate marketplace can ship a
-  same-named plugin (`ai-ios-engineering-toolkit:create-ui-component` vs the public
-  `create-component`) and dispatch still resolves correctly — the dual-name
+  same-named plugin (`ai-ios-toolkit:create-ui-component` vs the public
+  `create-component`) and dispatch still resolves correctly - the dual-name
   fallback lives in the pipeline, the implementation stays entirely in whichever
   plugin is enabled.
 - **Pipeline ↔ dev-toolkit boundary = MCP protocol, versioned contract.** The
@@ -233,5 +233,5 @@ that step's own contract).
   contract.
 - **Neither boundary is symmetric.** The pipeline depends on both other repos being
   present *for specific task types* (component work, UI testing) but functions
-  without either — a non-component bugfix task never touches the plugin marketplace,
+  without either - a non-component bugfix task never touches the plugin marketplace,
   and a task with no UI-testing step never opens the MCP connection.

package/docs/features.md

@@ -45,14 +45,14 @@ Build commands, test runners, lint tools, and review focus areas all adapt to th
 
 ### Stack Selection (marketplace plugins)
 
-Stack skill sets ship as versioned plugins in the `multi-agent-plugins` marketplace. Selecting a stack enables the matching plugin(s) in the current repo's `.claude/settings.json` `enabledPlugins` - no skill copying, no session restart tricks, no directory shuffling. The `ai-common-engineering-toolkit` (accessibility audit, humanizer, Firebase) is always enabled alongside the stack plugin.
+Stack skill sets ship as versioned plugins in the `multi-agent-plugins` marketplace. Selecting a stack enables the matching plugin(s) in the current repo's `.claude/settings.json` `enabledPlugins` - no skill copying, no session restart tricks, no directory shuffling. The `ai-common-toolkit` (accessibility audit, humanizer, Firebase) is always enabled alongside the stack plugin.
 
 ```bash
-/multi-agent:stack ios         # ai-ios-engineering-toolkit (SwiftUI, Xcode, HIG)
-/multi-agent:stack android     # ai-android-engineering-toolkit (Compose, Gradle, Hilt)
+/multi-agent:stack ios         # ai-ios-toolkit (SwiftUI, Xcode, HIG)
+/multi-agent:stack android     # ai-android-toolkit (Compose, Gradle, Hilt)
 /multi-agent:stack mobile      # iOS + Android combined
 /multi-agent:stack backend     # ai-backend-toolkit (spec-driven APIs)
-/multi-agent:stack frontend    # ai-frontend-engineering-toolkit (React/TSX)
+/multi-agent:stack frontend    # ai-frontend-toolkit (React/TSX)
 /multi-agent:stack fullstack   # backend + frontend
 /multi-agent:stack all         # every stack plugin
 ```
@@ -273,7 +273,7 @@ Turn a recurring, project-specific job into a first-class `/multi-agent:<name>`
 
 ### Figma / Component Generation (dispatched to marketplace plugins)
 
-Component + Figma-to-code work is no longer bundled in this repo. When Phase 0 classifies a task as `component`, Phase 3 dispatches it to the per-stack marketplace plugins (`ai-ios-engineering-toolkit` / `ai-android-engineering-toolkit` in the `multi-agent-plugins` marketplace) via the Skill tool. The plugin's component skill generates `{Name}Configuration.swift`, `{Name}View.swift`, `{Name}+Modifiers.swift`, `{Name}.figma.swift`, and `FIGMA.md` with a variant matrix, then runs a 14-item pre-commit checklist covering design tokens, accessibility, tests, and Code Connect.
+Component + Figma-to-code work is no longer bundled in this repo. When Phase 0 classifies a task as `component`, Phase 3 dispatches it to the per-stack marketplace plugins (`ai-ios-toolkit` / `ai-android-toolkit` in the `multi-agent-plugins` marketplace) via the Skill tool. The plugin's component skill generates `{Name}Configuration.swift`, `{Name}View.swift`, `{Name}+Modifiers.swift`, `{Name}.figma.swift`, and `FIGMA.md` with a variant matrix, then runs a 14-item pre-commit checklist covering design tokens, accessibility, tests, and Code Connect.
 
 The plugin's cross-cutting integration skills feed component detection + implementation when the design triggers them (content: form / price / ui-patterns; interaction: navigation / overlays / bottom-sheets). Each is native-SwiftUI-first and reads project specifics (token namespaces, component paths, UI systems) from `figma-config`, including the optional `ui.navigationSystem` / `ui.overlaySystem` / `ui.sheetSystem` hooks (absent -> stock SwiftUI), so the same capabilities work on any SwiftUI codebase. The plugin's evolve-component skill reconciles an existing component against current Figma (drift-heal) and additively extends it, behind a human gate.
 

package/index.js

@@ -59,7 +59,9 @@ if (command === "--version" || command === "-v" || command === "version") {
     npx @mmerterden/multi-agent-pipeline uninstall --yes      Skip prompt
     npx @mmerterden/multi-agent-pipeline uninstall --dry-run  Report what would be removed
     npx @mmerterden/multi-agent-pipeline uninstall --claude   Only Claude Code
+    npx @mmerterden/multi-agent-pipeline uninstall --copilot  Only Copilot CLI
     npx @mmerterden/multi-agent-pipeline uninstall --codex    Only Codex CLI
+    npx @mmerterden/multi-agent-pipeline uninstall --all-data ALSO remove pipeline settings + logs/state/metrics
     npx @mmerterden/multi-agent-pipeline uninstall --cursor   Legacy pre-v10.7 adapter-file cleanup (also --copilot-chat / --antigravity; --target=<path> overrides cwd)
 
   Help:

package/install/_codex-agents.mjs

@@ -24,8 +24,17 @@ import { ensureDir, ensureRealDir, isDryRun, writeFile } from "./_common.mjs";
  * The pipeline's ladder is `fable -> opus -> sonnet -> haiku`. Codex offers no
  * Anthropic models, so each tier maps onto an OpenAI model plus an effort
  * setting: effort carries the depth distinction that the model id carries on
- * Claude Code. Keep this table in sync with the Codex column of the Phase 4
- * reviewer matrix and with `pipeline/scripts/cost-table.json`.
+ * Claude Code.
+ *
+ * This table resolves a persona's DEFAULT tier, which is not the same thing as
+ * a per-slot override. Phase 4 dispatches Reviewer 3 with an explicit
+ * `gpt-5.6` @ `medium` (see `phases/phase-4-review.md`, `claude-md-template.md`
+ * and `reviewer-output.schema.json`, which all state that value) even though
+ * the persona's own tier is `sonnet` and resolves here to `gpt-5.4`. That is
+ * deliberate: the Codex panel buys its diversity from effort, so two slots
+ * share a model at different efforts while a third changes model. Do not
+ * "reconcile" the two by editing either side without deciding which behavior
+ * you want, and keep this table in sync with `pipeline/scripts/cost-table.json`.
  */
 export const CODEX_TIER_MAP = Object.freeze({
   fable: { model: "gpt-5.6", reasoning_effort: "xhigh" },