refactor(doctrine): C2 coherence — 30 doctrine/skill tensions resolved, doctrine wins (BDR-099)

One ask policy; mandated executors exempt from the delegation rule; skill plan satisfies the planning rule; journal line exempt from the approval gate; chore = maintenance without new behaviour; small fix on develop = bugfix; BDR-068 written as the one auto-finish exception; deploy routes to /deploy. Skills and agents follow: hotfix types by base + skips the design gate on trivial; capitalize/close create missing registries; commit-change asks the branch type; doc/seo/web-validate/refactor branch through the aiguillage; tour reports BREAKING fixes as needs-decision and runs doc-syncer two-mode; client-handover applies audit bundles from its main loop behind one gate; init-project/onboard use the 200-file graphify signal and bootstrap memory; release-candidate gates the tag push only; push wording aligned with the BDR-095 hooks; stale pointers fixed (§ Language, .gsd/ROADMAP.md, handover script path, design-gate lists).

CLAUDE.global.md

@@ -34,14 +34,16 @@ Apply unless repo-specific instructions override.
 2. Either missing → create it first (templates: `~/.claude/templates/memory/`).
 
 ## Workflow
-- Confirm before implementing only when real trade-offs exist (several
+- Confirm the approach only when real trade-offs exist (several valid
-  valid approaches, breaking change, destructive action); else proceed.
+  approaches, breaking change, destructive action); ask on the visible,
-  Minimal changes unless a broader refactor is requested. State trade-offs.
+  public-name and open-scope choices below; otherwise proceed. Minimal
+  changes unless a broader refactor is requested. State trade-offs.
 - Sub-agents: one task each, main context stays clean. Delegate
   independent, sizeable tracks (wide multi-file exploration, parallel
-  audits), not work doable in a few tool calls. Skill-mandated gates (fresh
+  audits), not work doable in a few tool calls. Skill-mandated executors and
-  verifier/security/challenge) always dispatch as written; a failed gate
+  gates (pinned executors, fresh verifier/security/challenge) always dispatch
-  re-dispatches a fresh executor, never redo its work by hand. A brief never
+  as written, whatever the task size; a failed gate re-dispatches a fresh
+  executor, never redo its work by hand. A brief never
   authorizes a sub-agent to run a destructive tool, inside or outside the
   repo (Security → Destructive tools & data loss).
 - Ask rather than guess. A choice visible in the result (placement,
@@ -67,7 +69,9 @@ Apply unless repo-specific instructions override.
 - Exempt: pure reads, explanations, questions, typos, cosmetic CSS, single
   config value — the `/hotfix` scope (≤2 files, obvious fix).
 - Once it qualifies: plan before code → one subtask = one coherent change
-  → check off as you go → high-level note at each milestone.
+  → check off as you go → high-level note at each milestone. A plan a skill
+  persists (`.claude/tasks/plans/`, contract) satisfies the rule; TODO.md
+  then carries one line per run.
 
 ## After code changes
 1. Run tests, lint, build, type-check if available. Report what was
@@ -98,12 +102,13 @@ exact. Pattern `[thing] [action] [reason]. [next step].` Registries load
 every session; caveman cuts ~40% of the tokens with no substance lost.
 Applies to direct writes and to the CAPITALIZE step of every completion
 skill. Prompts to the user may mirror their language; the entry is English.
-Legacy entries: compress on demand.
+Legacy entries: compress on demand via `/prune-memory`.
 
 **Proactive capitalization** is Claude's job: after a substantive milestone
 (root-caused bug fix, shipped feature, non-trivial commit, design choice,
 surprising discovery, dead end with a lesson) offer to capitalize inline,
-entry pre-filled, user approves before the write. Completion skills
+entry pre-filled, user approves before the write; the one-line journal
+entry is exempt, it logs and decides nothing. Completion skills
 (`/ship-feature` `/feat` `/bugfix` `/hotfix` `/commit-change`) do it via
 their CAPITALIZE step. Session close (`/close` = `/capitalize --ritual`):
 what was decided → decisions, learned → learnings, blocked → blockers.
@@ -137,20 +142,24 @@ never bent to match a newer one.
 ## Version control — gitflow (universal)
 Every git action follows gitflow, inside a skill or for an ad-hoc commit.
 `main` (prod) · `develop` (integration, off main) · `feature/*` `bugfix/*`
-`chore/*` (off develop → develop; chore = memory/doc maintenance such as a
+`chore/*` (off develop → develop; chore = maintenance without new behaviour:
-standalone `/capitalize` `/close` `/prune-memory` `/reconcile`) ·
+memory, docs, cleanup, `/refactor`, `/tour` fixes) · `release/*` (off develop
-`release/*` (off develop → main + back-merge develop) · `hotfix/*` (off main
+→ main + back-merge develop) · `hotfix/*` (off main → main + develop + any
-→ main + develop + any open release). `master` → `main` everywhere.
+open release; prod incidents only, a small fix on develop is a `bugfix`).
+`master` → `main` everywhere.
 
 Never commit code on `main` or `develop`: branch first as `<type>/<name>`
-(`.claude/**` memory/config commits are hook-exempt, following the work).
+(a `.claude/**` memory/config commit that follows merged work is hook-exempt
+and lands in place; a standalone memory task branches `chore/*`).
 Branch, merge and delete only via the lib: `bash ~/.claude/lib/gitflow.sh
 start <type> <name>` · `finish` · `delete <br>`. `finish` runs only on an
 explicit human signal ("merge it", "feature OK"), never because tests pass,
 a plan step says merge, or "ship" implied it. Assistance flows (`/feat`
 `/bugfix` `/hotfix`) and the standalone memory/doc skills auto-branch on a
-protected base but commit in place on a working branch, never finishing, so
+protected base but commit in place on a working branch, never finishing
-they branch to `chore/*` via the aiguillage, not the `.claude/**` exemption.
+(one exception, BDR-068: `/capitalize` and `/close` auto-finish the
+memory-only `chore/*` they created this run), so they branch to `chore/*`
+via the aiguillage, not the `.claude/**` exemption.
 Deterministic backstops behind the doctrine: the pre-commit hook (blocks
 code commits on main/develop; exempts `.claude/**`, `.githooks/**`, merges,
 the root commit), Gitea branch protection on both, and never `--no-verify`.
@@ -229,7 +238,8 @@ cryptic names.
   gates, registries). investigate only on explicit ask for the gstack
   ecosystem (cross-project learnings, /freeze, long open-ended investigation)
 - feat / hotfix / bugfix distinguished by file count → see descriptions
-- Ship / deploy / PR → ship (ship-feature if gstack off)
+- Ship / PR → ship (ship-feature if gstack off); deploy → deploy (runbook,
+  the user runs it)
 - Docs post-ship → document-release (doc if gstack off); stale-doc audit → doc
 - Grouped all-axes sweep ("tir groupé", fix + loop until clean) → tour
 - Open-work inventory / "queue empty?" / stale TODO vs git → reconcile

USAGE.md

@@ -262,7 +262,7 @@ cd mon-projet-existant/
 | 2  | Config baseline (onboarder agent) | CLAUDE.md, settings.json, .claudeignore, .claude/tasks/ + .claude/memory/ + .claude/audits/ |
 | 3  | Interview deep = business minimum (users, deadlines, équipe, légal, perfs) + adaptative par archétype | brief enrichi |
 | 3.5| ctx7 doc audit — fast-libs détectées, cache pré-fetché si besoin | .ctx7-cache/ |
-| 4  | Graphify (si complexity ≥ 30%) | graphify-out/GRAPH_REPORT.md |
+| 4  | Graphify (proposé dès 200 fichiers code, l'utilisateur décide) | graphify-out/GRAPH_REPORT.md |
 | 5  | Analyze read-only (analyzer agent) | .onboard-audit/analyze.md |
 | 6  | Audits parallèles selon archétype : | .onboard-audit/*.md (9 fichiers max) |
 |    |   — dette tech (general-purpose, audit read-only) |
@@ -280,7 +280,6 @@ cd mon-projet-existant/
 ```
 /onboard "Python FastAPI"              # hint stack
 /onboard force-archetype:wordpress     # override detection
-/onboard add gsd                       # générer ROADMAP.md pour GSD v2 (seul)
 ```
 
 **Après /onboard :**
@@ -291,6 +290,8 @@ cat .claude/audits/ONBOARD_REPORT.md
 # Démarrer la première tâche P0 avec le skill recommandé
 # (indiqué dans .claude/tasks/TODO.md)
 /hotfix "<titre P0>"   # ou /feat, /ship-feature, /bugfix selon le cas
+
+# Multi-session (GSD) : gsd init à la main — voir docs gsd-pi
 ```
 
 **Archétypes supportés (P1)** : static-html, wordpress, nextjs-app-router, astro-static, react-spa, rest-api-node, rest-api-python, cli-tool, library, dotfiles-meta.

agents/client-handover-writer.md

@@ -61,7 +61,7 @@ and degrading Google's NAP-consistency signal.
 
 Pipeline (each step gates the next):
 1. Baseline audits: SEO+GEO and security hardening in parallel.
-2. Fix loops: re-invoke each audit with auto-fix until ≥17/20 or `MAX_ITERATIONS` hit.
+2. Fix loops: apply each audit's bundle (AUTO items, ONE gate for GATED ones) and re-audit until ≥17/20 or `MAX_ITERATIONS` hit.
 3. Commit + push if files changed.
 4. Deploy pause: list deploy artifacts + process, wait for user confirmation.
 5. Live-site validation against the deployed URL.
@@ -266,14 +266,14 @@ For web projects, dispatch in **a single message with two parallel Agent calls**
 
 | Audit (web)   | Subagent          | Prompt template |
 |---------------|-------------------|-----------------|
-| SEO + GEO     | `general-purpose` | "Read `~/.claude/skills/seo/SKILL.md` and execute it on this project. The /seo skill runs SEO + GEO in parallel and writes a unified report to `.claude/audits/SEO.md`. Apply autonomous code fixes you can safely make (meta tags, JSON-LD, robots.txt, sitemap.xml, llms.txt, alt attrs, canonical tags). At the top of the report, the /seo skill MUST emit two distinct labeled score lines (already specified in its SKILL.md §1): `Score SEO (classique) : X.X / 20` and `Score GEO (IA) : X.X / 20`, plus the weighted global. The handover orchestrator parses SEO and GEO separately, so do not collapse them into a single `Score:` line. Return when the report file is written." |
+| SEO + GEO     | `general-purpose` | "Read `~/.claude/skills/seo/SKILL.md` and execute it on this project. The /seo skill runs SEO + GEO in parallel and writes a unified report to `.claude/audits/SEO.md`. Run it in conservative mode: apply NOTHING, return the FIX BUNDLE (the handover main loop applies it — see 'Applying the bundle'). At the top of the report, the /seo skill MUST emit two distinct labeled score lines (already specified in its SKILL.md §1): `Score SEO (classique) : X.X / 20` and `Score GEO (IA) : X.X / 20`, plus the weighted global. The handover orchestrator parses SEO and GEO separately, so do not collapse them into a single `Score:` line. Return when the report file is written." |
-| HARDEN        | `general-purpose` | "Read `~/.claude/skills/harden/SKILL.md` and execute it on this project. Apply autonomous code fixes (security headers in vercel.json/netlify.toml/.htaccess/nginx.conf, HSTS, CSP defaults, HTTP→HTTPS redirects, canonical, 404 page). Write report to `.claude/audits/HARDEN.md` with `Score: X/20` (or `X/100`) at the top. Return when the report file is written." |
+| HARDEN        | `general-purpose` | "Read `~/.claude/skills/harden/SKILL.md` and execute it on this project with `--fix` up to READY TO APPLY only: prepare the FIX BUNDLE (security headers, HSTS, CSP, HTTP→HTTPS redirects, canonical, 404 page), apply NOTHING — the handover main loop gates and applies it (see 'Applying the bundle'). Write report to `.claude/audits/HARDEN.md` with `Score: X/20` (or `X/100`) at the top. Return when the report file is written." |
 
 Non-web variant:
 
 | Audit (non-web) | Subagent          | Prompt template |
 |-----------------|-------------------|-----------------|
-| CSO             | `general-purpose` | "Read `~/.claude/skills/cso/SKILL.md` and execute in **daily mode** (8/10 confidence gate). Apply autonomous fixes for findings that are clearly safe (e.g., adding `.env` to `.gitignore`, replacing committed example secrets with placeholders). Write report to `.claude/audits/CSO.md` with `Score: X/20` (or `X/100`) at the top." |
+| CSO             | `general-purpose` | "Read `~/.claude/skills/cso/SKILL.md` and execute in **daily mode** (8/10 confidence gate). Apply NOTHING: return the fixes as a FIX BUNDLE (gitignore additions and placeholder swaps tagged AUTO, dependency upgrades tagged GATED — the handover main loop applies AUTO and gates the rest, see 'Applying the bundle'). Write report to `.claude/audits/CSO.md` with `Score: X/20` (or `X/100`) at the top." |
 
 Wait for both subagents to complete (parallel return).
 
@@ -370,7 +370,12 @@ iteration = 1
 # HARDEN/CSO/VALIDATE loops use only their own score.
 while (audit == "SEO" ? (SCORE_SEO < 17 OR SCORE_GEO < 17) : score < 17) \
       and iteration ≤ MAX_ITERATIONS:
-    re-dispatch the audit subagent with iteration context (see prompt below)
+    apply the pending FIX BUNDLE from THIS main loop — AUTO items, then
+        ONE gate for the GATED items (see "Applying the bundle" below);
+        iteration 1 consumes the baseline audit's bundle, if any
+    re-dispatch the audit subagent in AUDIT mode with iteration context
+        (see prompt below) — it re-scores and returns the next FIX BUNDLE;
+        it applies NOTHING (a dispatched child cannot hold a gate)
     re-parse score(s) AND projected code-only score(s) from the audit file
     if no scores improved AND no files changed → break (no progress)
     # Code-ceiling break: when the actual score has caught up with the
@@ -386,11 +391,43 @@ The projected code-only scores come from the analyzers' mandatory
 console). If no projected line is parseable, treat projected = 17
 (legacy behavior: loop chases 17 blindly).
 
+### Applying the bundle (THIS main loop — the child never applies)
+
+A dispatched child cannot hold a gate (harden: "the fix mode prepares the
+bundle; the dispatcher confirms"; geo: "NEVER apply a GATED item before
+explicit approval"). So every bundle is applied from here, per iteration:
+
+1. **AUTO items** — the tier the audit itself classed no-confirmation
+   (`/seo` STEP 1.5: seo batches A/B/C, geo G1–G4/G6). Dispatch each item's
+   L1 applier (`hotfixer` / `feater`) serially, item pasted verbatim,
+   "Do NOT commit — apply and self-verify only". Harden has no AUTO tier:
+   its whole bundle is confirmation-gated. CSO (non-web): gitignore additions
+   and secret placeholder swaps are AUTO.
+2. **GATED items** — seo D/E, geo G5, EVERY harden item (CSP, redirects,
+   anything its STEP 2b challenge flagged "could BREAK the site") and CSO
+   dependency upgrades (a version bump can break the build). Collect
+   them from every bundle of this iteration and present ONE gate
+   (AskUserQuestion): change → impact → file. Apply only the approved items
+   (same appliers); declined ones stay in the report as CODE-BLOQUÉ. Never
+   apply a GATED item before explicit approval.
+3. **USER ACTIONS** (seo batch F, geo G7) — never applied; they bound the
+   projected code-only score.
+
+For each item applied, append a line to the audit's fix log
+(`.claude/audits/SEO-FIX-LOG.md` for SEO and GEO items,
+`.claude/audits/HARDEN-FIX-LOG.md` for harden, `.claude/audits/CSO-FIX-LOG.md`
+for CSO) in the format
+`iter<N>: [SEO|GEO|HARDEN] <issue> → <file:line> — <action>`.
+
 ### Re-dispatch prompt template (SEO + GEO loop)
 
 Send to `general-purpose` subagent (`model: "fable"`):
 
-> Read `~/.claude/skills/seo/SKILL.md` and re-run it on this project.
+> Read `~/.claude/skills/seo/SKILL.md` and re-run it on this project in
+> **conservative (audit-only) intervention mode**: re-score and leave both
+> FIX BUNDLEs in `.claude/audits/SEO.md` ready-to-apply. Apply NOTHING and
+> ask the user NOTHING — the handover main loop is the dispatcher: it
+> applies the AUTO items and holds the gate on the GATED ones.
 > Previous scores:
 > - **SEO classique: `<SCORE_SEO_PREVIOUS>`/20** (threshold 17/20 — `<PASS|FAIL>`)
 > - **GEO (IA): `<SCORE_GEO_PREVIOUS>`/20** (threshold 17/20 — `<PASS|FAIL>`)
@@ -398,34 +435,33 @@ Send to `general-purpose` subagent (`model: "fable"`):
 > Iteration `<N>` of `<MAX_ITERATIONS>`. Both axes are gated independently;
 > the orchestrator continues to loop while EITHER score is below 17/20.
 >
-> Read `.claude/audits/SEO.md` for the current issue list. Apply ALL safe
+> Read `.claude/audits/SEO.md` for the current issue list. Prioritize
-> autonomous fixes (do not skip "easy" ones). Prioritize fixes for the
+> bundle items for the axis currently below threshold (do not drop "easy"
-> axis currently below threshold:
+> ones):
-> - SEO classique fixes: meta tags, headings, canonical, sitemap.xml,
+> - SEO classique: meta tags, headings, canonical, sitemap.xml,
 >   alt attrs, internal linking, Core Web Vitals hints.
-> - GEO (IA) fixes: llms.txt / llms-full.txt, robots.txt entries for AI
+> - GEO (IA): llms.txt / llms-full.txt, robots.txt entries for AI
 >   crawlers (GPTBot, ClaudeBot, PerplexityBot, etc.), Schema.org for AI
 >   extraction (QAPage, Speakable, Person+Article, HowTo, Organization
 >   graph), entity SEO (sameAs, @id), TL;DR / definition-lead content
 >   shape, citable stats markup, freshness signals.
 >
-> For each fix applied, append a line to `.claude/audits/SEO-FIX-LOG.md`
+> Update `.claude/audits/SEO.md` with the new scores — both labeled lines
-> (format: `iter<N>: [SEO|GEO] <issue> → <file:line> — <action>`). Update
+> MUST be present: `Score SEO (classique) : X.X / 20` and
-> `.claude/audits/SEO.md` with the new scores — both labeled lines MUST
+> `Score GEO (IA) : X.X / 20`, plus the weighted global.
-> be present: `Score SEO (classique) : X.X / 20` and
-> `Score GEO (IA) : X.X / 20`, plus the weighted global. Do NOT ask the
-> user; apply or skip with one-line justification in the fix log.
 
 ### Re-dispatch prompt template (HARDEN loop)
 
 Send to `general-purpose` subagent (`model: "fable"`):
 
-> Read `~/.claude/skills/harden/SKILL.md` and re-run it. Previous score:
+> Read `~/.claude/skills/harden/SKILL.md` and re-run it with `--fix` ONLY
+> up to the bundle: stop at `READY TO APPLY — awaiting dispatcher
+> confirmation`, apply NOTHING, ask NOTHING — the handover main loop is
+> the dispatcher that confirms. Previous score:
 > **`<SCORE_HARDEN_PREVIOUS>`/20** — below threshold. Iteration `<N>` of
-> `<MAX_ITERATIONS>`. Apply all autonomous fixes (security headers, HSTS,
+> `<MAX_ITERATIONS>`. The `## 8. Fix bundle` MUST cover security headers,
-> CSP, redirects, canonical, 404, .htaccess/nginx/vercel/netlify config).
+> HSTS, CSP, redirects, canonical, 404, .htaccess/nginx/vercel/netlify
-> Append entries to `.claude/audits/HARDEN-FIX-LOG.md`. Update
+> config. Update `.claude/audits/HARDEN.md` with the new score.
-> `.claude/audits/HARDEN.md` with new score.
 
 ### Re-dispatch prompt template (CSO loop — non-web only)
 
@@ -433,15 +469,18 @@ Send to `general-purpose` subagent (`model: "fable"`):
 
 > Read `~/.claude/skills/cso/SKILL.md` and re-run it in **daily mode**.
 > Previous score: **`<SCORE_CSO_PREVIOUS>`/20** — below threshold.
-> Iteration `<N>` of `<MAX_ITERATIONS>`. Apply all safe autonomous fixes
+> Iteration `<N>` of `<MAX_ITERATIONS>`. Apply NOTHING: return the fixes as
-> (gitignore additions, secret placeholder swaps, dependency upgrades for
+> a FIX BUNDLE in `.claude/audits/CSO.md` — gitignore additions and secret
-> known CVEs with semver-compatible patches). Append entries to
+> placeholder swaps tagged AUTO, dependency upgrades (even semver-compatible
-> `.claude/audits/CSO-FIX-LOG.md`. Update `.claude/audits/CSO.md` with
+> CVE patches) tagged GATED. The handover main loop applies AUTO items and
-> new score.
+> gates the rest (see 'Applying the bundle'). Update `.claude/audits/CSO.md`
+> with the new score.
 
 ### Parallelism
 
-For web projects, the two loops run in parallel: dispatch SEO iteration
+For web projects, the two loops run in parallel: apply both pending
+bundles first (AUTO items, then ONE gate for every GATED item of both),
+then dispatch SEO iteration
 `N` AND HARDEN iteration `N` in a single message with two `Agent` calls,
 wait for both, re-parse both scores, decide whether each loop continues,
 then dispatch iteration `N+1` for the audits still below threshold (in
@@ -454,7 +493,7 @@ nothing to parallelize).
 
 Track `score_history[audit] = [iteration → score]`. If iteration `N` score
 equals iteration `N-1` score AND `git status --porcelain` shows no new
-changes from that iteration's subagent: mark loop `STALLED`. Break.
+changes from that iteration's apply step: mark loop `STALLED`. Break.
 
 ### Escalation on cap or stall
 
@@ -627,7 +666,8 @@ Dispatch `general-purpose` subagent (`model: "fable"`):
 > Read `~/.claude/skills/web-validate/SKILL.md` and execute against the
 > deployed URL: `<DEPLOYED_URL>`. Audit W3C HTML validity (validator.nu),
 > W3C CSS validity (jigsaw.w3.org), WCAG 2.1 a11y (axe-core, pa11y).
-> Apply autonomous fixes ONLY in source code (the client controls deploy);
+> Apply NOTHING (the client controls deploy; the handover main loop applies
+> the returned FIX BUNDLE from source, AUTO items only): return the fix list,
 > document remaining issues. Write report to `.claude/audits/VALIDATE.md`
 > with `Score: X/20` (or `X/100`) at the top.
 

agents/commit-changer.md

@@ -42,10 +42,13 @@ approval gates live in the `/commit-change` dispatcher, not here.
 
 ### Phase 0: Gitflow aiguillage (before any commit)
 
-**Follow `$HOME/.claude/lib/gitflow-aiguillage.md` — your type = `chore`.**
+**Follow `$HOME/.claude/lib/gitflow-aiguillage.md` — your type = the `TYPE:`
-On `main`/`develop` it branches first (to `chore/<short-kebab-name>` derived
+line of the dispatch prompt (`feature` / `bugfix` / `chore`, chosen by the user
-from the pending work) so the commits never land directly on a protected
+in the dispatcher; never hardcode `chore`).** On `main`/`develop` it branches
-base; on a working branch it's a no-op (commit in place). Never `finish`,
+first (to `<TYPE>/<short-kebab-name>` derived from the pending work) so the
+commits never land directly on a protected base; a protected base with NO
+`TYPE:` in the prompt → do not branch, report it under EDGE CASES so the
+dispatcher asks. On a working branch it's a no-op (commit in place). Never `finish`,
 never `merge`, never `push` — this engine only commits. Branching itself is
 not a write of the pending changes, so it belongs in propose mode: by the
 time `MODE: apply` runs (a fresh dispatch), the branch already exists and
@@ -146,8 +149,9 @@ criteria as the standalone `/capitalize` flow:
   fix** (a pattern, a gotcha, a surprising API behaviour) → draft an entry
   for `.claude/memory/learnings.md` (LRN-XXX).
 
-**Language rule**: draft entries in English (see CLAUDE.md "Memory
+**Language rule**: draft entries in English AND caveman — fragments,
-registries" § Language) — the dispatcher's approval exchange may mirror the
+articles dropped, code/IDs/quoted errors verbatim (CLAUDE.md "Memory
+registries", Always English, always caveman) — the dispatcher's approval exchange may mirror the
 user's language, but what you draft here is what gets written verbatim in
 `MODE: apply` if approved unedited.
 
@@ -225,9 +229,9 @@ Otherwise:
    (`.claude/memory/decisions.md`, `blockers.md`, `learnings.md`) and
    update each file's `## Index` table. Add a one-line summary of the
    commit batch to today's heading in `.claude/memory/journal.md`.
-3. **Language rule**: written entries are ALWAYS in English regardless of
+3. **Language rule**: written entries are ALWAYS in English and caveman,
-   the language used in the dispatcher's approval exchange (CLAUDE.md
+   regardless of the language used in the dispatcher's approval exchange
-   "Memory registries" § Language).
+   (CLAUDE.md "Memory registries", Always English, always caveman).
 4. **Then commit the memory** — follow
    `$HOME/.claude/lib/capitalize-commit.md`: it surgically commits what
    was just written (`.claude/memory` + `.claude/tasks` only, never

agents/handover-doc-writer.md

@@ -130,7 +130,7 @@ concrete, no jargon. One short paragraph per idea.
    [§6.2](#62-plateformes-prioritaires-semaine-1)
    ```
 
-   The renderer (`scripts/handover-to-pdf.sh`) uses pandoc with
+   The renderer (`$HOME/.claude/skills/client-handover/scripts/handover-to-pdf.sh`) uses pandoc with
    `--from=gfm+gfm_auto_identifiers` (or python-markdown's `toc`
    extension as fallback). Both auto-generate heading IDs in the
    GitHub-style slug:

agents/release-executor.md

@@ -9,7 +9,7 @@ model: sonnet
 
 You execute the mechanical parts of a gitflow release. The `/release-candidate`
 dispatcher owns every judgment call — the version number, the "is it time to
-release" decision, and both pushes — and owns the human gate that sits BETWEEN
+release" decision, and the tag push — and owns the human gate that sits BETWEEN
 your two spans. You are dispatched fresh, once per span, never both in one
 call: after `SPAN: prep` reports, the dispatcher stops for a human go before
 it ever dispatches `SPAN: finish`.
@@ -76,12 +76,15 @@ actual branch; never finish whatever happens to be checked out.
    output verbatim; do not attempt to resolve it yourself.
 2. **Tag AFTER finish, on `main`** — never before:
    `git tag -a v<X.Y.Z> main -m "release <X.Y.Z>"` (annotated, so it lands on
-   main's release-merge commit).
+   main's release-merge commit). Finish has already pushed `main` and
+   `develop` through the lib's hooks (BDR-095); the tag stays local until
+   the dispatcher's tag-push gate.
 
 ### Forbidden in this span
-`git push` (any remote, any ref — the dispatcher owns the push gate),
+`git push` (any remote, any ref — `main`/`develop` ride the lib's hook
-deciding the version number, the when-to-release decision, attribution
+pushes during finish; the dispatcher owns the tag-push gate), deciding the
-trailers of any kind.
+version number, the when-to-release decision, attribution trailers of any
+kind.
 
 ---
 

lib/design-gate.md

@@ -17,11 +17,11 @@ Check BOTH the task description AND the filesystem:
 - Framework UI: `tailwind`, `styled-component`, `emotion`, `chakra`, `radix`, `shadcn`, `headless`
 
 **Filesystem signals** (quick check, no deep scan):
-- Target files have `.tsx`, `.jsx`, `.css`, `.scss`, `.less`, or `.module.css` extension
+- Target files have `.tsx`, `.jsx`, `.vue`, `.svelte`, `.astro`, `.css`, `.scss`, `.less`, or `.module.css` extension
 - `tailwind.config` or `postcss.config` present in project root
 - `tokens/`, `theme/`, or `design-system/` directory exists
 - Storybook config (`.storybook/`) present
-- Animation lib in `package.json` deps: `motion`, `motion-v`, `framer-motion` (legacy), `gsap`, `@gsap/react`, `lottie-react`, `react-spring`, `popmotion`, `@formkit/auto-animate`
+- Animation lib in `package.json` deps: any package `is_anim_lib_installed` recognizes (`lib/animation-lib-check.sh`, the single source)
 
 ## DECISION
 
@@ -40,7 +40,7 @@ and if not, point at ONE command — `/profile design`.
 Tier does NOT change WHAT gets checked. Every non-trivial design tier draws from
 the one `design` profile — so the gate checks that profile's **design-core
 tools** (the `# GATE-BLOCK:` allowlist in `design.profile`: ui-ux-pro-max,
-frontend-design, emil-design-eng, design-motion-principles, impeccable, design-html,
+frontend-design, emil-design-eng, design-motion-principles, design-html,
 design-review, design-consultation, the `21st` CLI and `21st-ui-build` — the
 canary for the whole 21st skill pack). The profile also bundles
 browser/plan/shotgun tooling and graphify for convenience; those never trip the
@@ -149,7 +149,7 @@ says so, because init has to happen in the agent chat, not in an installer.
 
 **Fires when BOTH hold** — else stay silent:
 
-1. impeccable is active (`skills/impeccable` present, i.e. it did not trip §3).
+1. impeccable symlink present under `skills/` (non-blocking external — not on the `# GATE-BLOCK:` list, so §3 never checks it).
 2. The project has no `PRODUCT.md` at its root.
 
 Evaluate it on the same path as §4: after the toolchain resolves, never on the

lib/gitflow-aiguillage.md

@@ -21,11 +21,14 @@ The caller passes its TYPE:
 |--------|------|------|
 | `/feat` | `feature` | develop |
 | `/bugfix` | `bugfix` | develop |
-| `/hotfix` | `hotfix` | main |
+| `/hotfix` | `hotfix` on main · `bugfix` on develop | main · develop |
+| `/seo` aggressive · `/web-validate --fix` | `feature` | develop |
 | `/capitalize` · `/close` · `/prune-memory` · `/reconcile` | `chore` | develop |
+| `/doc` · `/refactor` | `chore` | develop |
+| `/commit-change` | asks the user (`feature` / `bugfix` / `chore`) before `start` — a branch name is a public name | develop |
 
-The `chore` row = **standalone memory/doc work**: the registry / TODO / doc
+The `chore` rows = **standalone memory/doc/hygiene work**: the registry / TODO /
-reconciliation & curation skills, run OUTSIDE an assistance flow. Inside `/feat`
+doc reconciliation & curation skills (+ `/refactor`), run OUTSIDE an assistance flow. Inside `/feat`
 `/bugfix` `/hotfix` `/ship-feature` a working branch already exists (this check
 returns WORKING) and the memory commit rides it. The aiguillage only fires when
 such a skill is invoked directly on `main`/`develop` — i.e. memory IS the work,
@@ -39,6 +42,8 @@ develop + push) when THEY branched a `chore/*` off develop this run (BDR-068 —
 scoped [[LRN-069]] exception; see the capitalize skill's STEP 5C). `/prune-memory`
 + `/reconcile` stay fully human-gated: never run `gitflow finish` from them.
 
-Note: `hotfix` branches off **main** (prod) even when invoked from `develop` — that
+Note: a `hotfix/*` branch forks off **main** (prod) and fans out to main + develop
-is the gitflow definition of a hotfix. For a dev-scoped small fix, use `/bugfix`
+at finish — that is the gitflow definition of a hotfix. Invoked from `develop`,
-(branches off develop).
+`/hotfix` therefore starts a `bugfix/*` (off develop): a `hotfix/*` there would
+miss develop's code and later merge to prod. The small-fix routing is unchanged;
+only the branch type follows the base.

lib/verify-secure-loop.md

@@ -59,9 +59,9 @@ Parse its single `VERIFY — VERDICT:` line:
 
 - `CONFORME` → go to GATE 2. (First-pass conforme = no loop.)
 - `ECARTS(n)` → hand the dev the CONTRACT path + the exact `CRITERIA` gap
-  lines (NOT-MET / out-of-scope), nothing else. Inline dev fixes in place;
+  lines (NOT-MET / out-of-scope), nothing else: re-dispatch a FRESH executor
-  a dispatched dev is re-dispatched FRESH with those inputs only. Then
+  with those inputs only, never redo the fix by hand. Then re-run GATE 0 and
-  re-run GATE 0 and re-dispatch a FRESH verifier. Repeat.
+  re-dispatch a FRESH verifier. Repeat.
   **Max 3 conformity iterations** → STOP + human escalation with the
   CRITERIA table (the contract-vs-realized diff).
 - `ABANDONED(n)` → direct human gate, never a dev loop (a dev cannot close
@@ -88,8 +88,8 @@ stdout-only, no Write).
 Parse its single `SECURITY — VERDICT:` line:
 
 - `PASS` → done, proceed to commit.
-- `BLOCK(n)` → hand the dev the `BLOCKING` list + the CONTRACT path (inline
+- `BLOCK(n)` → hand the dev the `BLOCKING` list + the CONTRACT path
-  fix, or FRESH executor re-dispatch). Then re-run GATE 0, then
+  (re-dispatch a FRESH executor, never fix by hand). Then re-run GATE 0, then
   **re-verify the REQUEST first** (GATE 1, fresh verifier) — a security fix
   can drift the behavior — **then re-run GATE 2** (fresh auditor), in that
   order. **Max 3 security iterations** → STOP + human escalation with the

skills/bugfix/SKILL.md

@@ -272,7 +272,7 @@ A bugfix with an understood root cause is almost always worth one entry:
    ```
 4. Append approved entries + update the Index. Add a line to today's heading in `.claude/memory/journal.md`.
 
-**Language rule**: written entries are ALWAYS in English (see CLAUDE.md "Memory registries" § Language). The interactive gate may mirror the user's language; the appended entries must not.
+**Language rule**: written entries are ALWAYS English AND caveman — fragments, articles dropped, code/IDs/quoted errors verbatim — per CLAUDE.md "Memory registries" (Always English, always caveman). The interactive gate may mirror the user's language; the appended entries must not.
 
 If the bug was trivial and the root cause not transferable → skip with `CAPITALIZE: trivial, skip`.
 
@@ -287,7 +287,9 @@ hash, and no-ops if nothing was written.
 - No fix without understanding the root cause first (STEP 2/3).
 - Reflection (GATHER, INVESTIGATE, DIAGNOSIS, contract, loop decisions) NEVER
   leaves this main loop; execution NEVER stays in it — the executor is the
-  sonnet-pinned bugfixer subagent (BDR-066).
+  sonnet-pinned bugfixer subagent (BDR-066). A skill-mandated executor is exempt from the doctrine's "don't delegate
+  few-tool-call work" rule (CLAUDE.md "Workflow" names that exception:
+  skill-mandated dispatches run as written).
 - The executor is re-dispatched FRESH on every round-trip (NEED-DECISION,
   ECARTS, BLOCK) — feedback travels as contract path + named
   gaps/decisions, never as transcript.

skills/capitalize/SKILL.md

@@ -9,7 +9,7 @@ description: |
   Triggers: "capitalize", "before clear/compact", "flush memory", "don't
   lose this", "avant de clear/compact", "capitalise ce qui manque",
   "close", "fin de journée", "checkpoint memory".
-argument-hint: "[--ritual] [--no-push] (scans conversation + git + TODO against .claude/memory/; --ritual adds the 3-question reflection; --no-push holds memory on the chore branch instead of the default auto-merge+push)"
+argument-hint: "[--ritual] [--no-push] (scans conversation + git + TODO against .claude/memory/; --ritual adds the 3-question reflection; --no-push holds memory on chore/<name>: pushed to origin by the hooks, NOT merged (finish skipped), merge when ready; default = auto-finish into develop)"
 allowed-tools:
   - Read
   - Edit
@@ -65,17 +65,22 @@ ls .claude/memory/decisions.md .claude/memory/learnings.md \
 ls .claude/tasks/TODO.md 2>/dev/null
 ```
 
-- `.claude/memory/` missing entirely → print and STOP (do NOT create here —
+- `.claude/memory/` missing entirely → create it first (CLAUDE.md "Session
-  that is `/onboard` / `/init-project` responsibility):
+  start": either missing → create from the templates), then proceed:
-  ```
+  ```bash
-  ⚠️  .claude/memory/ absent. Lance `/onboard` (ou `/init-project`) pour créer
+  mkdir -p .claude/memory
-     les registres avant de capitaliser.
+  cp -n ~/.claude/templates/memory/{decisions,learnings,blockers,evals,journal}.md .claude/memory/
   ```
+  `/onboard` stays the fuller setup (CLAUDE.md, settings, audits) — this
+  bootstraps only the five registries capitalize writes to.
 - Some registry files missing → name them, create each from
   `~/.claude/templates/memory/<name>.md`, continue.
-- `.claude/tasks/TODO.md` missing → the TODO reconcile volet (STEP 2B) is
+- `.claude/tasks/TODO.md` missing → create a minimal one, then run the TODO
-  **skipped**. Do NOT create it (same posture as the registries). Registries
+  reconcile volet (STEP 2B) on it:
-  still run.
+  ```bash
+  mkdir -p .claude/tasks
+  printf '# TODO\n\n## %s\n' "$(date +%Y-%m-%d)" > .claude/tasks/TODO.md
+  ```
 
 ## STEP 1 — SCAN THE CONVERSATION
 
@@ -164,7 +169,7 @@ concatenated class names" entry.)
 
 ## STEP 2B — TODO RECONCILE (both modes)
 
-Runs only if `.claude/tasks/TODO.md` exists (STEP 0). Two passes.
+Runs on `.claude/tasks/TODO.md` (created minimal at STEP 0 when absent). Two passes.
 
 **PASS A — done-detection (TODO → reality).** Detection is free — a capable
 agent already spots the finished tasks from the STEP 1 git scan. The only rule
@@ -358,7 +363,7 @@ Then the closing line — pick by the STEP 5C persist result (`<mode>` = `Contex
 flushed` for pre-wipe, `Session closed` for ritual):
 
 - **auto-persisted (default — branched off develop, pushed)** → `✅ <mode> + persisted to origin/develop (<short>). Next session: read .claude/memory/ at startup.`
-- **--no-push (held on branch)** → `✅ <mode> + committed on chore/<name>, NOT pushed (--no-push). Merge + push when ready.`
+- **--no-push (held on branch)** → `✅ <mode> + committed on chore/<name> — pushed to origin by the hooks, NOT merged (--no-push: finish skipped). Merge when ready.`
 - **push failed after merge** → `✅ <mode> + merged to develop — ⚠️ push FAILED (<reason>); merged locally, push manually.`
 - **WORKING branch (rode a feature branch)** → `✅ <mode> + committed <mem_hash> on <branch>. Integrates when the branch merges.`
 - **commit skipped (rc 3)** → keep the ✅ on the WRITE but make the gap loud, never
@@ -380,7 +385,7 @@ manual commit (rc 3).
 - **Append-only.** Never overwrite or renumber existing registry entries.
 - **Caveman English** registry bodies, always English. **The TODO is plain
   prose, never caveman** — caveman is registries-only.
-- **TODO reconcile runs only if TODO.md exists.** Never create it (STEP 0).
+- **TODO reconcile always runs** — a missing TODO.md is created minimal at STEP 0.
 - **PASS A checks only on an unambiguous task↔code/commit map.** Partial /
   umbrella / vague → leave unchecked. Never on assumption.
 - **PASS B captures only explicit to-dos**, deduped — never invented or
@@ -398,7 +403,8 @@ manual commit (rc 3).
   WORKING branch (memory rides feature/bugfix) or rc 3 skips it. NEVER auto-finish
   a branch the run did not create.
 - **Skip trivial** for the 4 ID registries; journal excepted.
-- `.claude/memory/` missing → STOP at STEP 0, do not create the structure here.
+- `.claude/memory/` missing → STEP 0 creates the five registries from the
+  templates (doctrine: either missing → create first), then proceeds.
 
 ## Common mistakes
 
@@ -417,7 +423,7 @@ manual commit (rc 3).
 | Dumping an architecture directive as a TODO task | Route orientation/policy directives to decisions.md (BDR), not the TODO. |
 | Writing a ritual answer fresh without dedup | Ritual answers go through STEP 2 like any candidate; a dup shows its existing ID. |
 | French/English entry text | Prompt may be French; written registry entry is always English. |
-| Creating `.claude/memory/` or `.claude/tasks/TODO.md` when absent | Not this skill's job — registries STOP and point to `/onboard`; TODO volet is skipped. |
+| Stopping on a missing `.claude/memory/` or `.claude/tasks/TODO.md` | Doctrine says create first — STEP 0 bootstraps the five registries + a minimal TODO from the templates, then proceeds; `/onboard` is the fuller setup. |
 
 ## Red flags — STOP
 

skills/client-handover/SKILL.md

@@ -40,8 +40,8 @@ The agent runs a **ship-and-handover pipeline** with explicit gates:
 1. **PRE-FLIGHT** — Detect git repo, project root, language, project type, web sub-type, NAP signals, stack.
 2. **BASELINE AUDITS** — Run /seo (SEO+GEO) and /harden in parallel. Capture initial scores (`SCORE_SEO_BEFORE`, `SCORE_GEO_BEFORE`, `SCORE_HARDEN_BEFORE`).
 3. **FIX LOOPS (parallel, bounded)** — For each audit < 17/20:
-   - Re-invoke the audit subagent with explicit instruction to apply auto-fixes.
+   - Apply the pending FIX BUNDLE from the MAIN loop: items the audit classed AUTO directly, then every GATED item (CSP, redirects, anything harden marks "could BREAK the site") presented to the user at ONE gate before applying.
-   - Re-score.
+   - Re-invoke the audit subagent in audit mode: it re-scores and returns the next FIX BUNDLE; it applies nothing (a dispatched child cannot hold a gate).
    - Repeat up to `MAX_ITERATIONS` (default 5).
    - If still < 17/20 after cap → escalate to user with concrete remaining issues; user decides continue / stop / manual intervention.
 4. **COMMIT + PUSH** — If files changed during fix loops, run /commit-change (atomic logical commits) then `git push`.
@@ -57,7 +57,7 @@ The agent runs a **ship-and-handover pipeline** with explicit gates:
    - **§6 Détails techniques (pour les curieux)** — vulgarized BDR decisions, phases with technical detail, optional glossary (score table NOT here — promoted to §2).
    - **§7 Annexe — plateformes externes** (web/local-business only).
    - **§8 Annexe — build & déploiement** (only if requested).
-9. **RENDER** — Write `LIVRAISON.md` (fr) or `HANDOVER.md` (en) at project root, then run `scripts/handover-to-pdf.sh` to produce the matching branded `.html` (always) and `.pdf` (when a PDF engine is on the host: weasyprint > wkhtmltopdf > chromium). HTML/PDF use the ZenQuality cover page, green palette, Inter + Playfair Display typography, running header/footer with project name + page numbers.
+9. **RENDER** — Write `LIVRAISON.md` (fr) or `HANDOVER.md` (en) at project root, then run `$HOME/.claude/skills/client-handover/scripts/handover-to-pdf.sh` to produce the matching branded `.html` (always) and `.pdf` (when a PDF engine is on the host: weasyprint > wkhtmltopdf > chromium). HTML/PDF use the ZenQuality cover page, green palette, Inter + Playfair Display typography, running header/footer with project name + page numbers.
 
 Flags:
 - `--skip-fix-loop` — run baseline audits once, skip auto-fix iterations.

skills/close/SKILL.md

@@ -8,7 +8,7 @@ description: |
   (that is /prune-memory).
   Triggers: "close", "end session", "ferme la session", "session close",
   "checkpoint memory", "what did we learn", "retro rapide", "fin de journée".
-argument-hint: "[--no-push] (runs capitalize in ritual mode; --no-push holds memory on the chore branch instead of the default auto-merge+push)"
+argument-hint: "[--no-push] (runs capitalize in ritual mode; --no-push holds memory on chore/<name>: pushed to origin by the hooks, NOT merged (finish skipped), merge when ready; default = auto-finish into develop)"
 allowed-tools:
   - Read
   - Edit

skills/commit-change/SKILL.md

@@ -45,16 +45,28 @@ git config user.email
 - `git config user.email` empty → STOP, ask the user to configure identity
   first, do not dispatch.
 
-On a protected base (`main`/`develop`) the subagent runs the gitflow
+On a protected base (`main`/`develop` — `bash "$HOME/.claude/lib/gitflow.sh"
-aiguillage itself inside `MODE: propose` (its Phase 0) and branches to
+protected-base`) ask the user the branch TYPE before any dispatch — a branch
-`chore/*` before drafting the plan — code never lands directly on a
+name is a public name:
-protected branch.
+
+```
+AskUserQuestion:
+  Protected base — branch type for these commits? (feature / bugfix / chore)
+```
+
+Suggest `chore` only when every pending path is under `.claude/**` or docs;
+pending code never lands on a `chore/*` branch. Pass the answer as
+`TYPE: <type>` in the STEP 1 prompt — the subagent runs the aiguillage with it
+inside `MODE: propose` (its Phase 0) and branches to `<type>/*` before drafting
+the plan. Code never lands directly on a protected branch. On a working
+branch, omit `TYPE:` (the aiguillage is a no-op there).
 
 ## STEP 1 — Propose
 
 ```
 Agent(subagent_type="commit-changer", model="opus")
 prompt: "MODE: propose
+TYPE: <the STEP 0 answer — feature / bugfix / chore; omit on a working branch>
 $ARGUMENTS"
 ```
 
@@ -86,7 +98,7 @@ AskUserQuestion:
   BDR-077 — never redrawn inline on the session model); show the redrawn
   plan and re-ask.
 - `skip` → exit cleanly, no commits created, no `MODE: apply` dispatch.
-  Note: if the propose run created a `chore/*` branch (gitflow aiguillage
+  Note: if the propose run created a `<type>/*` branch (gitflow aiguillage
   off a protected base), that branch stays checked out with the work
   uncommitted — mention it so the user isn't surprised by the branch switch.
 

skills/deploy/SKILL.md

@@ -182,7 +182,6 @@ Author a runbook, seed the incident ledger, commit both, then proceed to STEP 1.
    #!/usr/bin/env bash
    # === deploy runbook (reference) — NOT run directly. Instantiated into the deploy checklist per delta. ===
    # Fixed steps run every deploy; annotated steps (@delta lines) re-instantiate from the delta.
-   # @config push_deploy_tags=false
    ```
 3. Scan for migration, rebuild, and dependency steps; propose `@delta:` annotations inline:
    - Migration steps (`psql -f`, `migrate up`, `supabase migration`) →
@@ -219,12 +218,10 @@ Author a runbook, seed the incident ledger, commit both, then proceed to STEP 1.
 | Backup command | "Backup command before migrations?" | `pg_dump "$DB" > ~/backups/pre-deploy-$(date +%F-%H%M).sql` |
 | Health-check URL | "Health-check URL (expects HTTP 200)?" | `https://$DEPLOY_HOST/health` |
 | Rollback note | "One-line rollback note (optional)?" | omit if blank |
-| Push deploy tags | "`push_deploy_tags`? (true / false)" | `false` |
 
 **Using** `~/.claude/templates/deploy/PROCEDURE.md` **as base, populate** fields from interview answers + detected artifacts:
 - Substitute `$DEPLOY_HOST` with the supplied host (keep literal `$DEPLOY_HOST` if none given).
 - Include only the annotated steps whose artifact was detected; keep all fixed steps.
-- Set `# @config push_deploy_tags=<answer>` in the header.
 - Append the rollback note as `# ROLLBACK: <note>` at the end if provided.
 
 → **[GATE]** below.
@@ -427,9 +424,7 @@ Then:
 
 The deploy succeeded. Lay the oracle and close out.
 
-1. Read `# @config push_deploy_tags=` from the `PROCEDURE.md` header (default
+1. Pick `date = today` (`YYYY-MM-DD`); if `deploy/<date>` exists, suffix `-N`.
-   `false`). Pick `date = today` (`YYYY-MM-DD`); if `deploy/<date>` exists, suffix
-   `-N`.
 2. Write `.claude/deploy/STATE.json` (overwrite):
    ```jsonc
    { "deployed_sha": "<PENDING.target_sha>", "deployed_at": "<now ISO-8601>",
@@ -438,9 +433,10 @@ The deploy succeeded. Lay the oracle and close out.
    **`deployed_sha` = `PENDING.target_sha`, NOT current HEAD** — HEAD may have
    moved during the gap; the bridge's target is the deployed truth.
 3. `git tag -a deploy/<date> <PENDING.target_sha> -m "<summary>"`.
-4. If `push_deploy_tags=true` → `git push origin deploy/<date>` — **best-effort,
+4. No separate tag push: the oracle commit (step 5) fires the gitflow
-   non-fatal**: a push failure logs a warning, never blocks the mark (the tag is a
+   post-commit hook, which pushes with `--follow-tags`, so `deploy/<date>`
-   bookmark; `STATE.json` is the oracle).
+   rides along (BDR-095). A hook `push FAILED` warning never blocks the mark
+   (the tag is a bookmark; `STATE.json` is the oracle).
 5. Commit the oracle:
    ```bash
    bash ~/.claude/lib/deploy-commit.sh commit "chore(deploy): mark <date> @ <short>" \

skills/doc/SKILL.md

@@ -22,6 +22,13 @@ Run the two-mode doc pipeline (BDR-077 — audit judgment on opus, patch on
 the sonnet pin, the validation gate in THIS loop; a dispatched agent cannot
 hold a gate):
 
+0. AIGUILLAGE — before any write, follow `$HOME/.claude/lib/gitflow-aiguillage.md`
+   — this skill's TYPE = `chore`. On `main`/`develop` it branches to
+   `chore/<name>` off develop so the doc patch lands on a branch; on a working
+   branch it proceeds in place. Never `gitflow finish`. `lib/doc-commit.md`'s
+   rc 5 (commit rejected by the hook on a protected base) stays as the
+   backstop, not the plan.
+
 1. AUDIT — dispatch:
    `Agent(subagent_type="doc-syncer", model="opus")`
    prompt: "MODE: audit. Audit public docs for this project. Context from

skills/feat/SKILL.md

@@ -244,7 +244,7 @@ Valider ? (all / <IDs> / edit / skip)
 
 Always append a 1-line entry to today's heading in `.claude/memory/journal.md`.
 
-**Language rule**: written entries are ALWAYS in English (see CLAUDE.md "Memory registries" § Language). The interactive gate may mirror the user's language; the appended entries must not.
+**Language rule**: written entries are ALWAYS English AND caveman — fragments, articles dropped, code/IDs/quoted errors verbatim — per CLAUDE.md "Memory registries" (Always English, always caveman). The interactive gate may mirror the user's language; the appended entries must not.
 
 If no substantive capture candidate → skip with `CAPITALIZE: nothing to log`.
 
@@ -259,7 +259,9 @@ hash, and no-ops if nothing was written.
 - Max 5 files. If more needed → `/ship-feature`.
 - Reflection (scope, plan, contract, loop decisions) NEVER leaves this main
   loop; execution NEVER stays in it — the executor is the sonnet-pinned
-  feater subagent (BDR-066).
+  feater subagent (BDR-066). A skill-mandated executor is exempt from the doctrine's "don't delegate
+  few-tool-call work" rule (CLAUDE.md "Workflow" names that exception:
+  skill-mandated dispatches run as written).
 - The executor is dispatched FRESH on every round-trip — feedback travels
   as contract path + named gaps/decisions, never as transcript.
 - Design gate only (not full plugin check). See STEP 0.5.

skills/gitflow/SKILL.md

@@ -85,7 +85,10 @@ On a protected base, assistance skills (`feat`/`bugfix`/`hotfix`) AND the standa
 memory/doc skills (`capitalize`/`close`/`prune-memory`/`reconcile`, TYPE `chore`)
 call `start <type>` to branch first; on a working branch they commit in place. Same
 `protected-base` predicate the out-of-skill hook uses. Caller→type map + rationale:
-`lib/gitflow-aiguillage.md`.
+`lib/gitflow-aiguillage.md`. `/capitalize` and `/close` auto-finish their memory-only
+`chore/*` branch into develop when THEY created it this run (BDR-068; `--no-push`
+opts out) — the only finish that fires without a live human signal; everything else
+stays human-gated.
 
 ## Failure modes (mechanical — lib return codes are the contract)
 

skills/hotfix/SKILL.md

@@ -64,8 +64,11 @@ disposition required at hotfix weight.
 
 Follow `$HOME/.claude/lib/design-gate.md`:
 - Scan $ARGUMENTS and target files for design/UI/style signals (CSS, component, styling, animation).
-- If signals found → run `design-tool-gate.sh`; if it reports INCOMPLETE,
+- Signals found → a hotfix is the trivial tier by definition (≤2 files, one
-  tell the user to run `/profile design` before proceeding.
+  cosmetic value — `design-gate.md` §1): skip the gate, no toolchain. If the
+  signals reveal real UI work (new component, layout, motion), this is not a
+  hotfix → route to `/feat` or `/bugfix` instead of running
+  `design-tool-gate.sh`.
 - If no signals → skip (zero overhead).
 
 ## STEP 1.7 — CONTRACT (silent autofill)
@@ -104,8 +107,12 @@ verify+secure loop).
 ## STEP 2 — PRE-FLIGHT
 
 **Gitflow aiguillage (before dispatch):** follow `$HOME/.claude/lib/gitflow-aiguillage.md`
-— your type = `hotfix`. On `main`/`develop` it branches first; on a working
+— your type follows the base: on `main` → `hotfix` (prod incident; finish fans
-branch it's a no-op (commit in place). Never `finish`.
+out to main + develop); on `develop` → `bugfix` (the fix forks from develop — a
+`hotfix/*` off main would miss develop's code and later merge to prod). Either
+protected base branches first; on a working branch it's a no-op (commit in
+place). The /hotfix size rules and the hotfixer executor are unchanged either
+way. Never `finish`.
 
 Snapshot current state so revert is possible:
 
@@ -223,7 +230,7 @@ Ask the user only when there is an actual candidate to propose.
 
 Always append a 1-line entry to today's heading in `.claude/memory/journal.md` (even trivial hotfix — journal is timeline, not signal).
 
-**Language rule**: the journal line and any proposed BLK/LRN entries are ALWAYS written in English (see CLAUDE.md "Memory registries" § Language).
+**Language rule**: the journal line and any proposed BLK/LRN entries are ALWAYS written English AND caveman — fragments, articles dropped, code/IDs/quoted errors verbatim — per CLAUDE.md "Memory registries" (Always English, always caveman).
 
 **Then commit the memory** — follow `$HOME/.claude/lib/capitalize-commit.md`: it
 surgically commits what capitalize just wrote (`.claude/memory` + `.claude/tasks`
@@ -237,7 +244,9 @@ trivial hotfix still produces a `chore(memory): journal — …` commit (Frame 2
 - Max 2 files changed. If more needed → `/bugfix`.
 - Reflection (LOCATE, contract, gate decisions) NEVER leaves this main
   loop; execution NEVER stays in it — the executor is the sonnet-pinned
-  hotfixer subagent (BDR-066).
+  hotfixer subagent (BDR-066). A skill-mandated executor is exempt from the doctrine's "don't delegate
+  few-tool-call work" rule (CLAUDE.md "Workflow" names that exception:
+  skill-mandated dispatches run as written).
 - The executor is dispatched FRESH, once — hotfix never re-dispatches after
   a failed or blocked attempt (it reverts and escalates to `/bugfix`, it
   does not retry). Sole exception: a class-tagged BLOCKED answered by the

skills/init-project/SKILL.md

@@ -103,7 +103,14 @@ every field the scaffolder consumes crosses the dispatch): BRIEF (verbatim)
 `~/.claude/CLAUDE.md`. A STOP (missing input) comes back as its report —
 resolve here, re-dispatch. The ~30s liveness pings are THIS loop's job
 while waiting.
-Creates: CLAUDE.md, `.claude/settings.json`, `.claudeignore`, `.gitignore`, `.env.example`, empty entry points. NO README, NO features, NO `.claude/tasks/` or `.claude/memory/` (not bootstrapped by this flow — copy from `~/.claude/templates/memory/` manually if wanted before STEP 10b's memory commit).
+Creates: CLAUDE.md, `.claude/settings.json`, `.claudeignore`, `.gitignore`, `.env.example`, empty entry points. NO README, NO features.
+Then bootstrap the memory in THIS loop, before STEP 5f so the root commit embeds
+it (doctrine: registries + TODO exist from day one; STEP 10b appends to them):
+```bash
+mkdir -p .claude/memory .claude/tasks
+cp -n ~/.claude/templates/memory/{decisions,learnings,blockers,evals,journal}.md .claude/memory/
+[ -f .claude/tasks/TODO.md ] || printf '# TODO\n\n## %s\n' "$(date +%Y-%m-%d)" > .claude/tasks/TODO.md
+```
 Verify: `git init` + build passes.
 
 ## STEP 5b — CREATE README
@@ -166,7 +173,7 @@ layout and the deterministic root commit:
 bash "$HOME/.claude/lib/gitflow.sh" init "chore: scaffold <project-name>"
 ```
 Creates `main`+`develop`, root-commits the FULL scaffold (CLAUDE.md, README,
-config, `.gitignore`, deps), reconciles the `.gitignore` socle, and installs the
+config, `.gitignore`, `.claude/memory/` + `.claude/tasks/TODO.md`, deps), reconciles the `.gitignore` socle, and installs the
 versioned pre-commit hook — all embedded in the root commit, working tree clean.
 This is the deterministic scaffold commit owner (closes BLK-010). The MVP is
 implemented on a `feature/*` branch off `develop` (STEP 8).
@@ -217,14 +224,15 @@ call. The plan is closed; execution and plan-conformity review are sonnet
 work. Reflection (task decomposition, review verdict arbitration) stays in
 this loop.
 
-## STEP 8b — GRAPHIFY FULL (after implementation)
+## STEP 8b — GRAPHIFY SIGNAL (after implementation — BDR-097)
-If `graphify` CLI is installed AND complexity >= 30%:
+graphify is proposed only from 200 tracked code files, and the USER decides —
-1. Run full graphify on the implemented project:
+never build, install or update a graph here:
-   ```bash
+```bash
-   graphify . --out graphify-out 2>/dev/null || true
+bash ~/.claude/lib/graphify-gate.sh .
-   ```
+```
-2. Print: `🔗 Full project graph updated at graphify-out/`
+- Prints a line → carry it into the FINAL OUTPUT status table as
-If `graphify` not installed or complexity < 30% → skip silently.
+  `GRAPHIFY: <line> — /graphify on your go`.
+- Silent → `GRAPHIFY: below 200 code files, not proposed`.
 
 ## STEP 9 — VERIFY + SECURE (fresh gates, bounded loops)
 Run the two fresh gates per `$HOME/.claude/lib/verify-secure-loop.md` with
@@ -284,7 +292,9 @@ capitalizes NOTHING. Do NOT fabricate a BDR to fill the step. Print
    [ decisions.md ]  BDR-XXX — <decision> — <1-line why>
    Valider lesquels ? (all / <IDs> / edit / skip)
    ```
-3. Append approved entries + update the Index. Append a journal line under today.
+3. Append approved entries to the existing registries (bootstrapped at STEP 5,
+   in the root commit) + update the Index. Append a journal line under today's
+   heading in `.claude/memory/journal.md`.
 
 **Hash rule — founding decisions carry NO commit hash; use path + date only.**
 This is by nature, not an omission: a founding decision is made at DESIGN
@@ -295,8 +305,10 @@ that IMPLEMENTS the decision, e.g. BDR-033 → 11792cc). This is the SECOND case
 where hash-anchoring does not apply — the first being a squash-merged PR, whose
 anchored commit ceases to exist.
 
-**Language rule**: written entries are ALWAYS in English (CLAUDE.md "Memory
+**Language rule**: written entries are ALWAYS English AND caveman — fragments,
-registries"). The gate may mirror the user's language; entries must not.
+articles dropped, code/IDs/quoted errors verbatim — per CLAUDE.md "Memory
+registries" (Always English, always caveman). The gate may mirror the user's
+language; entries must not.
 
 **Then commit the memory** — follow `$HOME/.claude/lib/capitalize-commit.md`: it
 surgically commits the approved founding decisions (`.claude/memory` +
@@ -369,5 +381,6 @@ LOCATION: <path> | STACK: <stack> | BUILD: ✅/❌ | TESTS: ✅<N>/❌
 V1 FEATURES: ✅<f> / ⚠️<f> partial: <reason>
 REMAINING ISSUES: <list or none>
 QUICK START: <exact cmds>
-CLAUDE.md ✅ | README ✅ | SETTINGS ✅
+CLAUDE.md ✅ | README ✅ | SETTINGS ✅ | MEMORY ✅
+GRAPHIFY: <STEP 8b line>
 ```

skills/onboard/SKILL.md

@@ -1,7 +1,7 @@
 ---
 name: onboard
 description: 'Use when bringing an existing repo into the claude-config framework — needs archetype detection, config install, full multi-axis audit (debt/SEO/GEO/UI-UX/perf/security/a11y/docs), and prioritized backlog. Multi-agent orchestrator. Do NOT use for repos created via /init-project. Triggers: "onboard", "onboard project", "audit existing repo", "setup existing project".'
-argument-hint: '[optional hints: "Python FastAPI" | "add gsd" | "Next.js monorepo" | "force-archetype:wordpress"]'
+argument-hint: '[optional hints: "Python FastAPI" | "Next.js monorepo" | "force-archetype:wordpress"]'
 allowed-tools: Read, Write, Edit, Bash, Glob, Grep, Agent, Skill
 ---
 
@@ -27,7 +27,8 @@ Run `$HOME/.claude/lib/plugin-gate.md` with hint "onboarding existing project +
 - PROPOSED CHANGES exist → show list, ask "Apply? (yes / no / customize)". Apply on confirm.
 - OK → `✅ Plugin check passed — [active plugins] — complexity: <score>%`, continue.
 
-Complexity score is carried forward for STEP 4 graphify decision.
+Complexity score is informative here; STEP 4 graphify is gated by
+`graphify-gate.sh` (200 tracked code files, BDR-097), not by this score.
 
 ---
 
@@ -107,7 +108,7 @@ L'agent génère :
 - `.claudeignore`
 - `.gitignore` (safety check)
 - `.claude/tasks/TODO.md`, `.claude/memory/{decisions,learnings,blockers,journal,evals}.md`
-- **Pas encore** `ROADMAP.md` (généré uniquement via `/onboard add gsd` — voir Next steps)
+- **Pas encore** `ROADMAP.md` (GSD multi-session : `gsd init` à la main, voir docs gsd-pi — cf. Next steps)
 
 Si `CLAUDE.md` existe déjà : lire son contenu, ne PAS écraser — fusionner après STEP 3.
 
@@ -151,11 +152,13 @@ Adopter le modèle gitflow sur ce repo existant :
 ```bash
 bash "$HOME/.claude/lib/gitflow.sh" init
 ```
-Sur un repo existant, cela : renomme `master`→`main` si besoin (LOCAL), crée
+Sur un repo existant, cela : renomme `master`→`main` si besoin (LOCAL), pose le
-`develop` depuis main, réconcilie le socle `.gitignore` (additif — n'écrase
+socle (`.gitignore` réconcilié — additif, n'écrase jamais les règles du projet —
-jamais les règles du projet), installe le hook pre-commit versionné, et fait UN
++ `.githooks/` versionnés) sur une branche `chore/gitflow-adopt` créée depuis
-commit `chore: adopt gitflow socle + pre-commit hook` sur main (pendant que le hook est
+main, la merge `--no-ff` dans main (le hook pre-commit est global sur la machine
-inactif → jamais auto-bloqué). Idempotent — un re-run est un no-op.
+et bloque tout commit de code sur main, mais exempte les merges — BDR-095),
+supprime la branche, puis crée `develop` depuis main. Idempotent — un re-run est
+un no-op.
 
 **Annoncer le renommage master→main** s'il a lieu. Le renommage est LOCAL ;
 repointer la branche par défaut du remote vers `main` + la protection de branche
@@ -247,21 +250,29 @@ Pour chaque fast-lib détectée :
 
 ---
 
-## STEP 4 — GRAPHIFY (si complexity ≥ 30% et pas déjà présent)
+## STEP 4 — GRAPHIFY (proposé dès 200 fichiers code — l'utilisateur décide)
 
 ```bash
 command -v graphify &>/dev/null && echo "available" || echo "not-installed"
-test -f graphify-out/GRAPH_REPORT.md && echo "graph-exists"
+test -f graphify-out/graph.json && echo "graph-exists"
+bash ~/.claude/lib/graphify-gate.sh .
 ```
 
 - **Pas installé** → skip avec message : `graphify non installé — skip audit architectural. Install : (voir graphify/SKILL.md)`
-- **Complexity < 30%** → skip silencieusement, projet trop petit pour justifier.
+- **`graph-exists`** → skip, réutiliser l'existant.
-- **Graphe déjà présent + récent** (fichier < 7j) → skip, réutiliser l'existant.
+- **La gate n'imprime rien** → moins de 200 fichiers code trackés (BDR-097 :
-- **Sinon** → run :
+  grep + lecture suffisent). Pas de proposition ; ligne FINAL OUTPUT
+  `below 200 code files, not proposed`.
+- **La gate imprime une ligne** (`graphify? N code files ≥ 200, no graph`) →
+  l'afficher et DEMANDER à l'utilisateur. Jamais de build sans oui explicite.
+  Sur oui :
   ```bash
-  graphify . --out graphify-out 2>&1 | tail -20
+  printf '.claude/\ndocs/superpowers/\n' >> .graphifyignore
+  grep -qxF 'graphify-out/' .gitignore || echo 'graphify-out/' >> .gitignore
+  graphify update .
   ```
-  Puis `test -f graphify-out/GRAPH_REPORT.md` pour valider.
+  Puis `test -f graphify-out/GRAPH_REPORT.md` pour valider. Sur non → skip,
+  ligne FINAL OUTPUT `proposed, declined`.
 
 Print : `🔗 Knowledge graph : graphify-out/GRAPH_REPORT.md (N nodes, M edges)`.
 
@@ -937,7 +948,7 @@ Choix ? (A / B / C / D / E)
 
 **STOP.** Attendre la réponse.
 
-- **A** → stop ici, l'utilisateur relira et reviendra avec `/onboard continue`.
+- **A** → stop ici, l'utilisateur relit les 4 fichiers ; le backlog (STEP 9) se génère ensuite à la demande depuis `.claude/audits/AUDIT_PROPOSALS.md`.
 - **B** → continuer STEP 9 avec toutes les recommandations.
 - **C** → demander les changements spécifiques, les appliquer dans .claude/audits/AUDIT_PROPOSALS.md, puis re-présenter la gate.
 - **D** → continuer STEP 9 avec seulement les P0.
@@ -1020,7 +1031,7 @@ Pour démarrer : lire .claude/tasks/TODO.md, choisir une tâche P0, lancer le /s
 - Si `CLAUDE.md` existe : le lire, ne pas l'écraser sans fusion après STEP 3.
 - STEP 3 : ne redemande jamais ce qui est déjà dans README ou manifests.
 - STEP 3.5 : si ctx7 absent + fast-libs, WARN mais ne bloque pas.
-- STEP 4 : skip si complexity < 30% ou graph récent déjà présent.
+- STEP 4 : graphify proposé seulement si `graphify-gate.sh` imprime une ligne (≥ 200 fichiers code, pas de graphe — BDR-097) ; l'utilisateur décide, jamais de build sans oui explicite.
 - STEP 5-6 : subagents isolés (Agent tool avec subagent_type spécifique) — pas de contexte partagé entre les audits. Chaque subagent écrit son rapport dans `.onboard-audit/<name>.md`.
 - STEP 6 dispatches parallélisables : regrouper dans un seul message Agent multi-calls.
 - `.onboard-audit/` gitignoré automatiquement — ne jamais commiter.
@@ -1035,7 +1046,7 @@ ARCHETYPE    : <name> (confiance: <niveau>)
 STACK        : <stack>
 CONFIG       : ✅ CLAUDE.md, settings.json, .claudeignore, .claude/{tasks,memory,audits}/
 CTX7 CACHE   : ✅ [libs] | ⚠️ not installed | — N/A
-GRAPHIFY     : ✅ graphify-out/ | ⚠️ not installed | — skipped (simple)
+GRAPHIFY     : ✅ graphify-out/ | ⚠️ not installed | — below 200 code files, not proposed | — proposed, declined
 AUDITS       :
   ✅ dette technique     (.onboard-audit/analyze.md + code-clean.md)
   ✅ sécurité            (.onboard-audit/cso.md)
@@ -1055,6 +1066,6 @@ SYNTHÈSE     :
 NEXT STEPS   :
   1. Ouvrir .claude/audits/ONBOARD_REPORT.md — overview complète
   2. Démarrer par la première tâche P0 de .claude/tasks/TODO.md avec le skill indiqué
-  3. /onboard add gsd — générer ROADMAP.md pour multi-session si besoin
+  3. GSD (multi-session) : `gsd init` à la main, voir docs gsd-pi
   4. .onboard-audit/ peut être supprimé (raw data consommée en synthèse)
 ```

skills/pdf-translate/SKILL.md

@@ -28,14 +28,17 @@ digraph pipeline {
 
 ## STEP 0: Dependencies
 
-Check before starting. Install what's missing.
+Check before starting. If something is missing, print the install
+command(s) and STOP until the user has run them — `sudo` / `apt` is the
+user's to run, never Claude's. The one exception: `pip install pymupdf`
+inside the project's own venv may be run by Claude.
 
 ```bash
-# Option A: poppler (lighter)
+# Option A: poppler (lighter) — USER runs the install
-command -v pdftoppm && echo "OK" || echo "INSTALL: sudo apt install poppler-utils"
+command -v pdftoppm && echo "OK" || echo "USER RUNS: sudo apt install poppler-utils"
 
 # Option B: PyMuPDF (more powerful — extracts embedded images with coordinates)
-python3 -c "import fitz; print('OK')" 2>/dev/null || echo "INSTALL: pip install pymupdf"
+python3 -c "import fitz; print('OK')" 2>/dev/null || echo "INSTALL: pip install pymupdf  (project venv only — otherwise USER RUNS)"
 ```
 
 Prefer PyMuPDF if both available — it extracts embedded images + gives page dimensions.

skills/refactor/SKILL.md

@@ -9,6 +9,12 @@ Dispatch the refactorer executor — behavior-preserving norm application is
 closed execution, so it runs pinned on **sonnet** (not the big session
 model). The scope you name is the only reflection; the agent applies norms.
 
+**Gitflow aiguillage first** (the refactorer edits code): follow
+`$HOME/.claude/lib/gitflow-aiguillage.md`, this skill's TYPE = `chore`. On
+`main`/`develop` run `bash ~/.claude/lib/gitflow.sh start chore refactor-<slug>`
+and dispatch on the new branch; on a working branch dispatch in place. Never
+`gitflow finish` — integration is human-gated.
+
 ```
 Agent(subagent_type="refactorer")
 prompt: "Refactor to strict project norms, preserving external behavior

skills/release-candidate/SKILL.md

@@ -24,7 +24,7 @@ The two mechanical spans (prep, finish+tag) run on the sonnet-pinned
 gate needed here, dispatch does the job. This dispatcher keeps everything
 the executor must never own: the version-NUMBER decision (judgment — derives
 from semver change nature), and the two human gates (when to release, and
-the push). A human gate sits BETWEEN the two spans by construction, so the
+the tag push). A human gate sits BETWEEN the two spans by construction, so the
 executor is never dispatched twice in one call.
 
 ## When to use
@@ -91,24 +91,26 @@ Parse the `RELEASE-EXEC REPORT`:
   the fan-out hit), STOP — resolving a conflicted fan-out is a human call,
   not an auto-retry.
 
-### STEP 6 — Push GATE (ASK)
+### STEP 6 — Tag push GATE (ASK)
-STOP. On explicit go only ([[LRN-069]]) — run the push HERE, in this
+`main` and `develop` are already on origin: the lib pushes every merge as
-dispatcher, never delegated to the executor:
+it lands (`_gitflow_merge_into` + the post-merge hook, BDR-095). Only the
+tag is left. STOP. On explicit go only ([[LRN-069]]) — run the tag push
+HERE, in this dispatcher, never delegated to the executor:
 ```
 AskUserQuestion:
-  Push main, develop, and v<X.Y.Z> to origin? — go / hold
+  Push tag v<X.Y.Z> to origin? — go / hold
 ```
 Go →
 ```bash
-git push origin main develop && git push origin v<X.Y.Z>
+git push origin v<X.Y.Z>
 ```
-`hold` → stop; the release is fanned out and tagged locally, unpushed.
+`hold` → stop; the release is on origin (main + develop), the tag stays local.
 
 ## Common mistakes
 - Tagging before `gitflow finish` → tag wouldn't sit on main's merge commit. Tag AFTER, on main.
 - Auto-firing finish because tests pass → finish is a HUMAN gate.
 - Restarting the tag at v1.0.0 → desyncs from the CHANGELOG lineage. Continue it.
-- Pushing without the ASK gate → [[LRN-069]].
+- Pushing the tag without the ASK gate → [[LRN-069]].
 
 ## Validation
 `RC_WORK=$(mktemp -d) RC_TAG=1 bash lib/tests/run-release-candidate.sh` → 5/5 (fan-out + tag on main). `RC_TAG=0` reds the tag assertion — proves the lib alone never tags (the gap this skill fills).

skills/seo/SKILL.md

@@ -535,6 +535,13 @@ intent, not header wording: **AUTO** = no-confirmation items (seo batches
 A/B/C · geo G1–G4/G6); **GATED** = items marked NEEDS CONFIRMATION / visible
 / structural (seo D/E · geo G5); **USER ACTIONS** = batch F / G7.
 
+### Gitflow aiguillage (before the first edit)
+
+Follow `$HOME/.claude/lib/gitflow-aiguillage.md` — this skill's TYPE =
+`feature` (aggressive mode edits code). On `main`/`develop` branch first:
+`bash ~/.claude/lib/gitflow.sh start feature seo-<slug>`; on a working
+branch apply in place. Never `gitflow finish` — integration is human-gated.
+
 ### Serial by ownership (no parallel race)
 
 The two bundles may touch the same shared template (meta vs JSON-LD). Apply

skills/ship-feature/SKILL.md

@@ -32,7 +32,8 @@ Verify the project has a `CLAUDE.md` and print a brief orientation summary:
 ls CLAUDE.md .claude/CLAUDE.md 2>/dev/null | head -1
 git branch --show-current 2>/dev/null || echo "not a git repo"
 git log --oneline -3 --format="%h %<(50,trunc)%s" 2>/dev/null || true
-ls .gsd/ROADMAP.md 2>/dev/null | head -1
+# gsd-pi ≥ 3: state in .gsd/STATE.md + gsd.db + milestones/<ID>/<ID>-ROADMAP.md (no .gsd/ROADMAP.md)
+ls .gsd/ 2>/dev/null >/dev/null && head -20 .gsd/STATE.md 2>/dev/null
 ```
 - **CLAUDE.md found** → read it silently, then print orientation header (informational, not a gate):
   ```
@@ -41,7 +42,7 @@ ls .gsd/ROADMAP.md 2>/dev/null | head -1
   Stack   : <stack from CLAUDE.md>
   Branch  : <current git branch>
   Recent  : <last 3 commit messages>
-  GSD     : <current milestone if .gsd/ROADMAP.md exists, else "not initialized">
+  GSD     : <current milestone read from .gsd/STATE.md if .gsd/ exists, else "not initialized">
   ```
   Continue to STEP 1.
 - **Not found** →
@@ -262,7 +263,7 @@ Feature shipped implies at least one design decision worth capturing. Run this B
 4. Append approved entries to the registries. Update the Index table at the top of each file.
 5. Append a one-line entry to `.claude/memory/journal.md` under today's date heading (`## YYYY-MM-DD`).
 
-**Language rule**: written entries are ALWAYS in English (see CLAUDE.md "Memory registries" § Language). The interactive gate above may mirror the user's language; the appended entries must not.
+**Language rule**: written entries are ALWAYS English AND caveman — fragments, articles dropped, code/IDs/quoted errors verbatim — per CLAUDE.md "Memory registries" (Always English, always caveman). The interactive gate above may mirror the user's language; the appended entries must not.
 
 If nothing substantive to log → print `CAPITALIZE: nothing substantive to log` and skip.
 

skills/tour/SKILL.md

@@ -39,7 +39,7 @@ plus its report IS the approval gate, reviewed by the human afterwards.
 Core principle: **autonomy on the working branch, never on shared
 state.** The skill may edit code freely on its own branch; it may NOT
 silently rewrite declared state (target TODO, memory registries) or
-integrate anything (merge/finish/push).
+integrate anything (merge/finish/push to `main`/`develop`).
 
 ## When NOT to use
 
@@ -79,8 +79,8 @@ Model discipline (the user-fixed invariant behind this mode):
   must never be pinned down to an executor tier.
 - Inside a runner, every dispatched agent keeps the tier this skill
   already defines: security-auditor (sonnet frontmatter), the Phase B
-  audit (analyzer opus pin or `model="opus"`), doc-syncer (sonnet
+  audit (analyzer opus pin or `model="opus"`), doc-syncer (audit on
-  frontmatter, its two-mode contract untouched).
+  `model="opus"`, patch on its sonnet frontmatter — BDR-077).
 
 Runner dispatch, one per project:
 
@@ -91,7 +91,7 @@ Agent(subagent_type="general-purpose",
     for EXACTLY ONE project: <absolute path>. Flags: <--report-only|none>.
     Skip STEP 0/0b (routing) and the global summary — the dispatcher owns
     them. Every rule of the skill applies unchanged: max 3 iterations,
-    never merge/finish/push, scoped commits, report appended to that
+    never merge/finish/push main|develop, scoped commits, report appended to that
     project's own .claude/audits/TOUR.md. Return EXACTLY: the project's
     one-line global-summary row (STEP 3 format), then BRANCH: <name|no
     branch>, then REPORT: <path>.")
@@ -156,9 +156,12 @@ honestly in the summary. Never loop past 3.
    MEDIUM/LOW → fix only if local and behavior-preserving, else leave
    `open`. Every fix minimal, CLAUDE.md security defaults apply.
    A CRITICAL/HIGH fix that changes the API contract (new required
-   header/param, changed status codes, moved paths) is still applied —
+   header/param, changed status codes, moved paths) is NOT applied — a
-   but its report row and the global summary line carry a **BREAKING**
+   breaking change is the human's call (CLAUDE.md: confirm before a
-   tag, so the human review cannot miss it.
+   breaking change). Its report row becomes
+   `open — needs decision (BREAKING)` with the proposed patch attached,
+   and the global summary line carries the **BREAKING** count.
+   Behaviour-preserving CRITICAL/HIGH fixes stay auto-applied.
 4. Commit scoped: `git add <files touched>` (never `-A`),
    `fix(security): …`.
 
@@ -196,11 +199,18 @@ honestly in the summary. Never loop past 3.
    `.claude/memory/`** — an inferred checkbox is exactly the lie
    /reconcile exists to catch. The human applies suggestions via
    `/reconcile` later.
-2. **Doc sync** — dispatch doc-syncer in AUTOMATIC (silent) mode:
+2. **Doc sync** — two-mode doc-syncer, mirrors /ship-feature STEP 8
-   public docs only (README, INSTALL, USAGE, CHANGELOG…), never
+   (BDR-077: audit judgment on opus, patch on the sonnet pin):
-   `.claude/**`, never CLAUDE.md. Commit its `PATCHED_FILES:` via
+   `Agent(subagent_type="doc-syncer", model="opus")` with `MODE: audit`
-   `bash ~/.claude/lib/doc-commit.sh` when available, else a scoped
+   + `auto-mode scope: <files this tour touched>`; public docs only
-   `docs: …` commit of exactly those paths.
+   (README, INSTALL, USAGE, CHANGELOG…), never `.claude/**`, never
+   CLAUDE.md. NONE → done. `[MINOR]` PATCH PLAN → re-dispatch
+   `Agent(subagent_type="doc-syncer")` (sonnet frontmatter) with
+   `MODE: patch` + the plan verbatim, then commit its `PATCHED_FILES:`
+   via `bash ~/.claude/lib/doc-commit.sh` when available, else a scoped
+   `docs: …` commit of exactly those paths. SIGNIFICANT → not applied:
+   report row `suggested` with the plan item (the tour has no human
+   gate mid-run).
 
 ### End of iteration
 
@@ -231,15 +241,15 @@ order:
 | ID | Axis | File | Sev | Finding | Status |
 |----|------|------|-----|---------|--------|
 | SEC-1 | security | app.py:17 | high | shell=True + concat | fixed |
-| SEC-2 | security | app.py:14 | high | no authz on POST /backup | fixed — **BREAKING**: new required X-Backup-Token header |
+| SEC-2 | security | app.py:14 | high | no authz on POST /backup | open — needs decision (BREAKING): new required X-Backup-Token header, patch attached |
 | CLN-1 | clean | utils.py:9 | - | dead legacy_md5 | fixed |
 | REC-1 | reconcile | TODO.md | - | "/health" unchecked, shipped 2d92696 | suggested |
 | DOC-1 | doc | README.md | - | phantom /status endpoint | fixed |
-Checks: pytest PASS, ruff PASS. Residuals: none. Commits: 5. BREAKING: 1 (SEC-2).
+Checks: pytest PASS, ruff PASS. Residuals: SEC-2 (needs decision). Commits: 4. BREAKING: 1 (SEC-2).
 ```
 
 Global summary inline, one line per project (append `BREAKING: n` to
-any project line whose fixes changed an API contract):
+any project line with contract-changing fixes left open for decision):
 
 ```
 TOUR COMPLETE — 2026-07-04
@@ -257,7 +267,10 @@ without that approval — neither this repo's nor any target project's.
 ## Rules
 
 - Branch via the gitflow lib; **never `gitflow finish`, never merge,
-  never push** — no exceptions, "the tour is green" is not a signal.
+  never push `main`/`develop`** — "the tour is green" is not a signal.
+  The chore branch's own commits are pushed by the gitflow hooks
+  (BDR-095); a `push FAILED` hook warning is a report residual, fixed
+  with a plain `git push -u origin chore/tour-<date>`.
 - Scoped pathspecs only; `git add -A` is forbidden.
 - Target TODO.md and target `.claude/memory/` are READ-ONLY. Reconcile
   produces suggestions, not edits.
@@ -286,12 +299,12 @@ without that approval — neither this repo's nor any target project's.
 | One TOUR.md for all projects in the config repo | Each project gets its own `.claude/audits/TOUR.md`. |
 | Fixing a behavior-changing "cleanup" finding | That is a bug → BUGS-FOUND.md, untouched code. |
 | Scratch audit files left untracked at the end | Delete them in STEP 3.2 — a dirty tree self-blocks the next tour. |
-| Contract-changing security fix reported as plain "fixed" | Tag **BREAKING** in the row AND the summary line. |
+| Contract-changing security fix auto-applied | Not applied: row `open — needs decision (BREAKING)` + proposed patch; **BREAKING** count in the summary line. |
 
 ## Red flags — STOP
 
 - About to `Edit` a target project's TODO.md or `.claude/memory/*`.
-- About to run `gitflow finish`, `git merge`, or `git push`.
+- About to run `gitflow finish`, `git merge`, or push `main`/`develop`.
 - About to `git add -A` or commit on `main`/`develop`.
 - Starting iteration 4, or "just one more loop, it's almost clean".
 - Security phase done without the security-auditor agent and without a

skills/web-validate/SKILL.md

@@ -299,7 +299,12 @@ Options :
   D) Abort — keep .claude/audits/VALIDATE.md as audit report
 ```
 
-4. On `A` : dispatch each file-group's applier at L1 (execution = sonnet;
+4. On `A` : gitflow aiguillage FIRST — follow
+   `$HOME/.claude/lib/gitflow-aiguillage.md`, this skill's TYPE = `feature`
+   (`--fix` edits code). On `main`/`develop` branch before any edit:
+   `bash ~/.claude/lib/gitflow.sh start feature web-validate-<slug>`; on a
+   working branch apply in place. Never `gitflow finish` (human-gated).
+   Then dispatch each file-group's applier at L1 (execution = sonnet;
    this loop only orchestrates), serially — one applier at a time, appliers
    share files:
 

templates/deploy/PROCEDURE.md

@@ -1,7 +1,6 @@
 #!/usr/bin/env bash
 # === deploy runbook (reference) — NOT run directly. Instantiated into the deploy checklist per delta. ===
 # Fixed steps run every deploy; # @delta: steps re-instantiate from the delta.
-# @config push_deploy_tags=false
 # NOTE grammar: glob=<pat>:each repeats the command per matching file (e.g. psql -f <each>);
 #               glob=<pat>:list runs once + lists matching files as VERIFY items; when=<pat,...> is conditional.
 # Style: one command per line, as typed in an interactive session — step 1 opens