@event4u/agent-config 9.17.0 → 9.18.0

package/dist/agent-src/contexts/execution/plan-confidence-gate.md

@@ -0,0 +1,133 @@
+# Plan-Confidence Gate (Gate C)
+
+Shared definition of the plan-confidence gate. Loaded by the gated
+plan-authoring surfaces — [`/roadmap:create`](../../commands/roadmap/create.md),
+the [`roadmap-writing`](../../skills/roadmap-writing/SKILL.md) skill,
+[`/feature:plan`](../../commands/feature/plan.md), and
+[`/feature:roadmap`](../../commands/feature/roadmap.md) — each of which links
+here instead of restating the gate. Machine-checked grammars (the C→R1 state
+schema, marker lines) are owned by
+[`docs/contracts/plan-review-gates.md`](../../../docs/contracts/plan-review-gates.md);
+this context carries the agent-side procedure.
+
+## Activation
+
+Read `planning.challenge_on_create` from `.agent-settings.yml` once per task
+and cache it. `true` or **missing key** → the gate is active. `false` → the
+gate is inert: proceed with the surface's normal flow, no marker line, no
+interview.
+
+**Explicit user bypass wins for that turn** — bypass phrasings and their
+equivalents drop the gate immediately
+(`DE: "einfach schreiben", "Interview überspringen" · EN: "just write it", "skip the interview"`).
+A bypass is **counted**, never punished: append a `gate_c_bypass` event to
+`agents/evidence/metrics/gate-metrics.jsonl` (see
+[`plan-review-gates § metrics`](../../../docs/contracts/plan-review-gates.md))
+so gate erosion stays visible.
+
+## When it fires — and when NOT
+
+**Fires** on a plan-artifact **authoring** ask reaching a gated surface: a new
+roadmap, feature plan, or feature roadmap, or a from-scratch rewrite of one.
+
+**Does NOT fire on:**
+
+- checkbox flips, dashboard regens, archival moves, inbound-ref rewrites;
+- prose-only or typo edits to an existing plan;
+- `/roadmap:process-*` execution runs (execution is not authoring);
+- artifacts that are not plans (ADRs, tickets, skills, rules).
+
+## Procedure
+
+### 1. Codebase first (challenge-me Step 0)
+
+Before assessing confidence, run the lookup: existing routes, conventions,
+prior similar features, the seed's cited files. **Never ask what `grep`,
+`view`, or `codebase-retrieval` answers in seconds** — repo findings resolve
+branches silently.
+
+### 2. Seed-time confidence assessment
+
+The measure is the four 95%-conditions from
+[`/challenge-me vision`](../../commands/challenge-me/vision.md) § "The four
+95% conditions". At seed time (before any interview turn) the check degrades
+to: **does the ask already pin (a) the goal, (b) the scope boundary
+(in/out), (c) the hard constraints, and (d) observable acceptance criteria —
+without guessing?** Any gap → below 95%.
+
+### 3a. Confident path
+
+All four hold → emit **exactly one** marker line, then proceed with the
+surface's normal flow:
+
+```
+> Confidence ≥ 95% — creating directly (planning.challenge_on_create)
+```
+
+No second marker, no summary of the assessment — one line, then the work.
+
+### 3b. Uncertain path
+
+Any gap → route into the [`/challenge-me vision`](../../commands/challenge-me/vision.md)
+interview with the plan ask as seed. On reaching 95% (or the user firing
+`!pitch`), the pitch feeds the surface's normal authoring flow — the same
+mechanic as the existing `!roadmap` routing. The interview replaces the
+surface's own clarification questions; it does not stack on top of them.
+
+### 3c. Inline degrade protocol (hosts without `/challenge-me`)
+
+`/challenge-me` ships in the `product-reasoning` pack; the gate never
+hard-depends on it. When the command is unavailable, run the interview
+inline with the same stop conditions:
+
+1. **Resolve first** — derive goal / scope / constraints / AC from the ask
+   plus the codebase lookup (step 1); only genuinely open, load-bearing
+   branches survive.
+2. **One question per turn** — surface one load-bearing branch with a
+   recommended option (numbered options per `user-interaction`;
+   one-question-per-turn per `ask-when-uncertain`).
+3. **Repeat until** every surfaced branch is resolved AND no new
+   load-bearing branch appeared after the last answer AND the cached draft
+   summary (goal + in/out + constraints + AC) is stable for two turns.
+4. **Then author** via the surface's normal flow.
+
+### 4. C→R1 handoff — write the state file
+
+On completing the gate (either path that ran an interview), write
+`agents/runtime/state/gate-c-<plan-slug>.json` so the R1 risk pass never
+re-asks resolved branches — the user is never interviewed twice for one
+plan. Schema (canonical:
+[`plan-review-gates § C→R1 handoff`](../../../docs/contracts/plan-review-gates.md)):
+resolved branches, plan hash, timestamp, and a mandatory `transcript_ref`
+(path + content hash of the interview transcript artifact).
+
+**Write-path rule:** only the Gate C flow writes
+`agents/runtime/state/gate-c-*.json`. This binds the **agent**; there is no
+lint and no hook entry behind it, and the state file is gitignored so CI never
+sees it — a human reading the diff is what catches a violation (stated
+`enforced_by: none` in
+[`plan-review-gates § 4.1`](../../../docs/contracts/plan-review-gates.md)).
+The rule defends against silent agent shortcuts, not against the local human
+(who holds the settings escape hatch anyway).
+Confident-path runs (no interview) write no state file; R1 then runs its
+risk pass fresh.
+
+## Non-goals
+
+- **No execution authorization.** The gate ends where authoring ends — the
+  post-artifact hard stop (`scope-control § Authoring vs. implementation`)
+  is unchanged; no gate outcome licenses implementation.
+- **No safety-floor changes.** Hard Floor, commit policy, and git-ops gates
+  are untouched by any gate path.
+- **No firing on maintenance touches** (see "When it fires" above).
+
+## See also
+
+- [`/challenge-me vision`](../../commands/challenge-me/vision.md) — the four
+  95%-conditions and the interview this gate routes into.
+- [`plan-review-gates`](../../../docs/contracts/plan-review-gates.md) — machine
+  grammars: C→R1 state schema, Risk Register, completion review.
+- [`ask-when-uncertain`](../../rules/ask-when-uncertain.md) — the
+  one-question-per-turn floor the degrade protocol runs under.
+- [`scope-control`](../../rules/scope-control.md) — the post-artifact hard
+  stop this gate never softens.

package/dist/agent-src/rules/active-remediation.md

@@ -16,6 +16,8 @@ routes_to:
   - "guideline:agent-infra/active-remediation-mechanics"
 workspaces: [engineering]
 packs: [engineering-base]
+collision_ok:
+  "refactor": "an explicit clean-up/refactor ask is the remediation ladder's own trigger"
 ---
 
 # Active Remediation

package/dist/agent-src/rules/architecture.md

@@ -11,6 +11,8 @@ routes_to:
   - "skill:module-detect-on-the-fly"
 workspaces: [agent-config-maintainer, engineering]
 packs: [meta]
+collision_ok:
+  "controller": "thin-handler floor on controller work"
 ---
 
 # Architecture Rules

package/dist/agent-src/rules/augment-edit-discipline.md

@@ -17,6 +17,8 @@ validator_ignore:
     reason: "Rule scopes the portability gate to the uncondensed authoring tree."
 workspaces: [agent-config-maintainer]
 packs: [meta]
+collision_ok:
+  ".augment/": "portability + sync floors for the .augment tree"
 ---
 
 # Augment Edit Discipline

package/dist/agent-src/rules/brand-consistency.md

@@ -1,15 +1,8 @@
 ---
-type: "auto"
+type: "manual"
 tier: "2a"
-description: "Generated UI/copy/assets check against active brand tokens + voice profile — untraceable values are flagged off-brand"
-triggers:
-  - keyword: "brand"
-  - keyword: "on-brand"
-  - keyword: "off-brand"
-  - keyword: "brand voice"
-  - keyword: "brand tokens"
-  - phrase: "brand consistency"
-  - phrase: "on brand"
+description: "Merged into brand-source-of-truth (2026-08-04) — every emitted colour/type/spacing/voice choice traces to a brand token or voice rule"
+alwaysApply: false
 applies_to_user_types:
   - "creator"
   - "developer"
@@ -21,56 +14,11 @@ packs: [brand]
 
 # Brand Consistency
 
-Brand is the layer that *constrains* UI, copy, and assets — not a coat of paint
-applied after. This is a concrete Validation gate: emitted artifacts are checked
-against the **active brand tokens and voice profile**, and anything that cannot
-be traced back to one is flagged off-brand. It mirrors the design-intelligence
-principle that **audit findings outrank corpus suggestions** — the consumer's
-brand outranks any generated default.
+**Iron Law.** Every emitted colour, type, spacing, and voice choice traces to a
+brand token or a voice rule — a value that traces to neither is off-brand.
 
-## The Iron Law
-
-```
-EVERY EMITTED COLOUR, TYPE, SPACING, AND VOICE CHOICE TRACES TO A BRAND TOKEN
-OR A VOICE RULE. A VALUE THAT TRACES TO NEITHER IS OFF-BRAND.
-NEVER SHIP AN ASSET THE ACTIVE BRAND CANNOT ACCOUNT FOR.
-```
-
-## What this gates
-
-- **Colour / type / spacing** in generated UI or assets that is not one of the
-  active brand tokens (from `state.ui_design` or the consumer's `.tokens.json`).
-  A raw hex / font / px value with no token behind it is off-brand.
-- **Copy and microcopy** whose register contradicts the active voice profile
-  (e.g. playful copy under an authoritative `Ruler` voice).
-- **Generated brand assets** (logo, banner, social) emitted without injecting the
-  registered brand tokens — raw generation is allowed only when no brand layer
-  exists (graceful fallback), never as a silent override of one that does.
-
-## When it fires
-
-A brand layer is active for the run (a `.tokens.json` / `state.ui_design` brand
-profile or a confirmed `brand-strategy` / `brand-identity` constraint set) AND a
-UI / copy / asset is being emitted or reviewed.
-
-## When NOT to fire
-
-- No brand layer exists yet (greenfield) — there is nothing to validate against;
-  generation proceeds on the brief, and `brand-strategy` / `brand-identity`
-  define the tokens first.
-- The user explicitly scopes an exploration ("just show options") with no
-  commitment to an on-brand deliverable.
-- Non-brand surfaces: scripts, CLI output, internal tooling.
-
-## How to surface a violation
-
-Name the off-brand value, the token/voice rule it should have matched, and the
-fix — do not silently rewrite. Untraceable value → "off-brand: `<value>` has no
-brand token; nearest is `<token>`" so the human decides token-or-exception.
-
-## See also
-
-- [`brand-source-of-truth`](brand-source-of-truth.md) — establishes which brand profile is authoritative for the run.
-- [`brand-identity`](../skills/brand-identity/SKILL.md) — defines the tokens this gate validates against.
-- [`brand-to-tokens`](../skills/brand-to-tokens/SKILL.md) — emits the DTCG `.tokens.json` source of truth.
-- [`design-intelligence`](../skills/design-intelligence/SKILL.md) — the "audit findings outrank corpus" precedent this gate mirrors.
+Body merged into [`brand-source-of-truth`](brand-source-of-truth.md) § Iron
+Law 2 (rule hygiene, 2026-08-04): the two rules shared triggers (`brand
+tokens`, `brand voice`) and one subject — the consumer brand as the run's
+authority. This file stays as a pointer so inbound references keep resolving;
+it is `type: manual` (reference-only, no router emission).

package/dist/agent-src/rules/brand-source-of-truth.md

@@ -1,12 +1,17 @@
 ---
 type: "auto"
 tier: "2a"
-description: "Consumer brand tokens + voice profile are the run's source of truth — the corpus fills gaps, never overrides"
+description: "Consumer brand tokens + voice profile are the run's source of truth — the corpus fills gaps, never overrides; emitted values that trace to no token are flagged off-brand"
 triggers:
+  - keyword: "brand"
+  - keyword: "on-brand"
+  - keyword: "off-brand"
   - keyword: "brand tokens"
   - keyword: "brand voice"
   - keyword: "brand guide"
   - keyword: "brand profile"
+  - phrase: "brand consistency"
+  - phrase: "on brand"
   - phrase: "brand source of truth"
 applies_to_user_types:
   - "creator"
@@ -19,18 +24,35 @@ packs: [brand]
 
 # Brand Source of Truth
 
-A precedence rule, light by design. When a consumer already has a brand —
-registered tokens (`.tokens.json`), a voice profile, a brand guide — that brand
-is authoritative for the run. The curated `brand` corpus (archetypes, colour
-psychology, type principles) is a **gap-filler**, never an override.
+A precedence rule plus its validation gate (merged from the former
+`brand-consistency` rule, 2026-08-04 — the pair shared triggers and one
+subject). When a consumer already has a brand — registered tokens
+(`.tokens.json`), a voice profile, a brand guide — that brand is authoritative
+for the run. The curated `brand` corpus (archetypes, colour psychology, type
+principles) is a **gap-filler**, never an override.
 
-## The Iron Law
+Brand is the layer that *constrains* UI, copy, and assets — not a coat of paint
+applied after. The gate half is a concrete Validation gate: emitted artifacts
+are checked against the **active brand tokens and voice profile**, and anything
+that cannot be traced back to one is flagged off-brand. It mirrors the
+design-intelligence principle that **audit findings outrank corpus
+suggestions** — the consumer's brand outranks any generated default.
+
+## Iron Law 1 — precedence
 
 ```
 CONSUMER BRAND TOKENS AND VOICE WIN. THE CORPUS FILLS GAPS ONLY.
 NEVER OVERWRITE A REGISTERED BRAND VALUE WITH A CORPUS DEFAULT.
 ```
 
+## Iron Law 2 — the consistency gate
+
+```
+EVERY EMITTED COLOUR, TYPE, SPACING, AND VOICE CHOICE TRACES TO A BRAND TOKEN
+OR A VOICE RULE. A VALUE THAT TRACES TO NEITHER IS OFF-BRAND.
+NEVER SHIP AN ASSET THE ACTIVE BRAND CANNOT ACCOUNT FOR.
+```
+
 ## Precedence order
 
 1. **Consumer brand profile** — registered `.tokens.json`, voice profile, brand
@@ -39,16 +61,42 @@ NEVER OVERWRITE A REGISTERED BRAND VALUE WITH A CORPUS DEFAULT.
 3. **The brand corpus** — archetype / colour / type / messaging defaults, used
    only where 1 and 2 are silent, and always surfaced as corpus-sourced.
 
+## What the gate covers
+
+- **Colour / type / spacing** in generated UI or assets that is not one of the
+  active brand tokens (from `state.ui_design` or the consumer's `.tokens.json`).
+  A raw hex / font / px value with no token behind it is off-brand.
+- **Copy and microcopy** whose register contradicts the active voice profile
+  (e.g. playful copy under an authoritative `Ruler` voice).
+- **Generated brand assets** (logo, banner, social) emitted without injecting the
+  registered brand tokens — raw generation is allowed only when no brand layer
+  exists (graceful fallback), never as a silent override of one that does.
+
 ## When it fires
 
-Any brand decision while a consumer brand profile is present — to keep the
-corpus from quietly replacing a value the brand already defines.
+- **Precedence half:** any brand decision while a consumer brand profile is
+  present — to keep the corpus from quietly replacing a value the brand already
+  defines.
+- **Gate half:** a brand layer is active for the run (a `.tokens.json` /
+  `state.ui_design` brand profile or a confirmed `brand-strategy` /
+  `brand-identity` constraint set) AND a UI / copy / asset is being emitted or
+  reviewed.
 
 ## When NOT to fire
 
 - No consumer brand exists (greenfield) — the corpus is the only ground; use it,
-  marked as corpus-sourced, until `brand-identity` defines real tokens.
-- Non-brand surfaces.
+  marked as corpus-sourced, until `brand-identity` defines real tokens. There is
+  nothing to validate against; generation proceeds on the brief, and
+  `brand-strategy` / `brand-identity` define the tokens first.
+- The user explicitly scopes an exploration ("just show options") with no
+  commitment to an on-brand deliverable.
+- Non-brand surfaces: scripts, CLI output, internal tooling.
+
+## How to surface a violation
+
+Name the off-brand value, the token/voice rule it should have matched, and the
+fix — do not silently rewrite. Untraceable value → "off-brand: `<value>` has no
+brand token; nearest is `<token>`" so the human decides token-or-exception.
 
 ## Design register note
 
@@ -61,6 +109,9 @@ for the full Brand-mode vs Product-mode discriminator.
 
 ## See also
 
-- [`brand-consistency`](brand-consistency.md) — validates emitted artifacts against the authoritative brand profile this rule establishes.
+- [`brand-consistency`](brand-consistency.md) — pointer stub; its body lives here since the 2026-08-04 merge.
 - [`brand`](../skills/brand/SKILL.md) — the gap-fill corpus this rule subordinates to consumer brand.
+- [`brand-identity`](../skills/brand-identity/SKILL.md) — defines the tokens the gate validates against.
+- [`brand-to-tokens`](../skills/brand-to-tokens/SKILL.md) — emits the DTCG `.tokens.json` source of truth.
+- [`design-intelligence`](../skills/design-intelligence/SKILL.md) — the "audit findings outrank corpus" precedent the gate mirrors.
 - [`docs/guidelines/design-modes.md`](../docs/guidelines/design-modes.md) — brand vs product register discriminator and routing.

package/dist/agent-src/rules/broken-access-control.md

@@ -19,6 +19,10 @@ routes_to:
   - "skill:multi-tenancy"
 workspaces: [engineering]
 packs: [engineering-base]
+collision_ok:
+  "controller": "controllers return user data — the ownership/tenant check floor"
+  "endpoint": "every data-returning endpoint needs the three negative tests"
+  "tenant": "cross-tenant isolation is this rule's core subject"
 ---
 
 # Broken Access Control

package/dist/agent-src/rules/cli-output-handling.md

@@ -30,6 +30,11 @@ routes_to:
   - "skill:rtk-output-filtering"
 workspaces: [agent-config-maintainer, engineering]
 packs: [meta]
+collision_ok:
+  "phpstan": "phpstan output is verbose — rtk wrap applies"
+  "composer": "composer output is verbose — rtk wrap applies"
+  "docker": "docker output is verbose — rtk wrap applies"
+  "rector": "rector output is verbose — rtk wrap applies"
 ---
 
 # Cli Output Handling

package/dist/agent-src/rules/code-comment-discipline.md

@@ -17,6 +17,9 @@ routes_to:
   - "guideline:code-clarity"
 workspaces: [engineering]
 packs: [engineering-base]
+collision_ok:
+  "refactor": "refactors churn comments — the keep/drop discipline applies"
+  "implement": "fresh implementation is where over-commenting lands"
 ---
 
 # Code Comment Discipline

package/dist/agent-src/rules/commit-conventions.md

@@ -10,6 +10,8 @@ routes_to:
   - "skill:conventional-commits-writing"
 workspaces: [engineering]
 packs: [engineering-base]
+collision_ok:
+  "commit": "format floor for every commit message"
 ---
 
 # Commit Conventions

package/dist/agent-src/rules/cross-source-consistency.md

@@ -23,6 +23,9 @@ routes_to:
   - "guideline:agent-infra/cross-source-consistency-mechanics"
 workspaces: [engineering, product]
 packs: [engineering-base, product-basic]
+collision_ok:
+  "mockup": "a mockup is a second source — discrepancy check against the ticket text"
+  "screenshot": "a screenshot is a second source — discrepancy check"
 ---
 
 # Cross-Source Consistency

package/dist/agent-src/rules/design-fidelity.md

@@ -32,6 +32,8 @@ routes_to:
   - "guideline:design-fidelity-mechanics"
 workspaces: [engineering]
 packs: [engineering-base, frontend-design]
+collision_ok:
+  "mockup": "a provided mockup is the spec — 1:1 fidelity floor"
 ---
 
 # Design Fidelity

package/dist/agent-src/rules/doc-screenshot-hygiene.md

@@ -25,6 +25,8 @@ applies_to_user_types:
   - "all"
 workspaces: [agent-config-maintainer, construction, engineering, finance, founder, gtm, legal-review-prep, ops, product, small-business]
 packs: [meta]
+collision_ok:
+  "screenshot": "a screenshot shipping into docs is the PII/anonymization gate"
 ---
 
 # Doc-Screenshot Hygiene

package/dist/agent-src/rules/docker-commands.md

@@ -11,6 +11,10 @@ routes_to:
   - "skill:docker"
 workspaces: [engineering]
 packs: [engineering-base]
+collision_ok:
+  "artisan": "artisan runs inside the project container, never on the host"
+  "composer": "composer runs inside the project container"
+  "docker": "its own core subject"
 ---
 
 # Docker Commands

package/dist/agent-src/rules/domain-adoption-policy.md

@@ -16,6 +16,8 @@ routes_to:
   - "guideline:agent-infra/domain-adoption-gates"
 workspaces: [agent-config-maintainer]
 packs: [meta]
+collision_ok:
+  "src/skills/": "a skill in a fresh vertical is exactly the domain-gate surface"
 ---
 
 # Domain Adoption Policy

package/dist/agent-src/rules/domain-safety-disclaimer.md

@@ -11,9 +11,10 @@ triggers:
   - keyword: "symptoms"
   - keyword: "dosage"
   - keyword: "medication"
+  # `valuation` / `DCF` moved to finance-safety-floor (2026-08-04) — that
+  # pack floor already carried both; this rule keeps the advisory-shaped
+  # finance triggers the floor does not cover.
   - keyword: "investment memo"
-  - keyword: "valuation"
-  - keyword: "DCF"
   - keyword: "tax position"
   - keyword: "strategic recommendation"
   - keyword: "board memo"
@@ -39,6 +40,8 @@ applies_to_user_types:
   - "creator"
 workspaces: [agent-config-maintainer, construction, engineering, finance, founder, gtm, legal-review-prep, ops, product, small-business]
 packs: [meta]
+collision_ok:
+  "review this contract": "contract review is advisory-shaped — the not-legal-advice footer floor"
 ---
 
 # Domain Safety — Advisory Disclaimer

package/dist/agent-src/rules/engineering-safety-floor.md

@@ -26,6 +26,8 @@ routes_to:
   - "skill:threat-modeling"
 workspaces: [engineering]
 packs: [engineering-base]
+collision_ok:
+  "migration": "a prod migration is a hard-floor evidence surface (rollback path, blast radius)"
 ---
 
 # Engineering Safety Floor

package/dist/agent-src/rules/framework-neutrality-in-generic-skills.md

@@ -34,6 +34,15 @@ workspaces: [agent-config-maintainer]
 packs: [meta]
 enforced_by:
   - "validator:src/scripts/lint_framework_leakage.ts"
+collision_ok:
+  "src/skills/": "generic skills must not mandate a stack — fires on skill edits"
+  "src/rules/": "rule edits are checked for framework leakage"
+  "src/agent-src/commands/": "command edits are checked for framework leakage"
+  "artisan": "artisan named in a generic artifact is leakage — the lint surface"
+  "eloquent": "eloquent named in a generic artifact is leakage — the lint surface"
+  "formrequest": "FormRequest named in a generic artifact is leakage — the lint surface"
+  "phpstan": "phpstan named in a generic artifact is leakage — the lint surface"
+  "rector": "rector named in a generic artifact is leakage — the lint surface"
 ---
 
 # framework-neutrality-in-generic-skills

package/dist/agent-src/rules/image-likeness-and-rights.md

@@ -18,6 +18,10 @@ applies_to_user_types:
 self_contained: true
 workspaces: [agent-config-maintainer, engineering, gtm]
 packs: [ai-image]
+collision_ok:
+  "/image:": "rights/likeness gate on the image-generation surface"
+  "likeness": "its own core subject — the rights gate"
+  "scripts/ai-image/adapters/": "adapter edits touch the likeness-gated generation path"
 ---
 
 # Image Likeness and Rights

package/dist/agent-src/rules/improve-before-implement.md

@@ -16,6 +16,10 @@ validator_ignore:
     reason: "Rule cites the authoring tree when describing where examples live."
 workspaces: [engineering]
 packs: [engineering-base]
+collision_ok:
+  "migration": "a migration is a significant behavioural change — the pre-implementation checks apply"
+  "refactor": "refactor = significant change; the demand gate + three checks apply"
+  "implement": "its own core trigger — pre-implementation validation"
 ---
 
 # Improve Before Implement

package/dist/agent-src/rules/laravel-routing.md

@@ -16,6 +16,10 @@ trust:
 install:
   default: false
   removable: true
+collision_ok:
+  "artisan": "artisan is a Laravel-flavoured PHP routing cue — load the laravel skill"
+  "eloquent": "eloquent is a Laravel routing cue"
+  "formrequest": "FormRequest is a Laravel routing cue"
 ---
 
 # Laravel Routing

package/dist/agent-src/rules/legal-safety-floor.md

@@ -28,6 +28,9 @@ packs: [legal-review-prep]
 trust:
   level: advisory
   human_review_required: true
+collision_ok:
+  "controller": "'controller' is a GDPR role term (controller/processor) in the legal pack"
+  "review this contract": "the legal-pack contract-review gate — both floors must fire"
 ---
 <!-- agent-config:human-review-banner -->
 > HUMAN REVIEW REQUIRED · trust: advisory · owner: unknown

package/dist/agent-src/rules/lethal-trifecta-guard.md

@@ -17,6 +17,9 @@ workspaces: [engineering]
 packs: [engineering-base]
 enforced_by:
   - "validator:src/scripts/lint_skill_frontmatter_safety.ts"
+collision_ok:
+  "src/skills/": "skill authoring reviews tool grants and the egress leg"
+  "src/agent-src/commands/": "command authoring reviews the three trifecta legs"
 ---
 
 # Lethal-Trifecta Guard

package/dist/agent-src/rules/media-governance-routing.md

@@ -31,6 +31,11 @@ workspaces: [agent-config-maintainer, engineering, gtm]
 packs: [ai-image, ai-video]
 enforced_by:
   - "validator:src/scripts/lint_media_policy_linkage.ts"
+collision_ok:
+  "/image:": "routes the project-local media policy layer on the same surface"
+  "/audio:": "routes the media policy layer for audio surfaces"
+  "/video:": "routes the media policy layer for video surfaces"
+  "likeness": "likeness prompts must load the policy layer before generation"
 ---
 
 # Media Governance Routing

package/dist/agent-src/rules/minimal-safe-diff.md

@@ -15,6 +15,8 @@ workspaces: [engineering]
 packs: [engineering-base]
 enforced_by:
   - "hook:minimal-safe-diff"
+collision_ok:
+  "fix": "a fix is the smallest change that solves the stated problem"
 ---
 
 # Minimal Safe Diff

package/dist/agent-src/rules/no-roadmap-references.md

@@ -20,6 +20,8 @@ packs: [meta]
 enforced_by:
   - "validator:src/scripts/check_no_roadmap_refs.ts"
   - "validator:src/scripts/check_council_references.ts"
+collision_ok:
+  "agents/roadmaps/": "roadmap edits are where transient-ref links get written"
 ---
 
 # No Transient References from Stable Artifacts

package/dist/agent-src/rules/notes-first-reasoning.md

@@ -15,6 +15,8 @@ routes_to:
   - "skill:memory-consolidation"
 workspaces: [agent-config-maintainer, engineering]
 packs: [meta]
+collision_ok:
+  "debug": "debug hypotheses belong in the session notes file"
 ---
 
 # Notes-First Reasoning

package/dist/agent-src/rules/php-coding.md

@@ -16,6 +16,8 @@ trust:
 install:
   default: false
   removable: true
+collision_ok:
+  "phpstan": "phpstan findings are the PHP pattern floor's own quality surface"
 ---
 
 # Php Coding

package/dist/agent-src/rules/prefer-enums-over-literals.md

@@ -12,6 +12,9 @@ triggers:
 self_contained: true
 workspaces: [engineering]
 packs: [engineering-base]
+collision_ok:
+  "migration": "migrations add columns — multi-state fields should land as enums from the start"
+  "category": "a category field is the canonical enum candidate"
 ---
 
 # Prefer Enums Over Literal Values

package/dist/agent-src/rules/provider-lifecycle-discipline.md

@@ -25,6 +25,11 @@ applies_to_user_types:
   - "maintainer"
 workspaces: [agent-config-maintainer, engineering, gtm]
 packs: [ai-image, ai-video]
+collision_ok:
+  "/image:": "provider-tier gate fires on the same generation surface"
+  "/audio:": "provider-tier gate for audio providers"
+  "/video:": "provider-tier gate for video providers"
+  "scripts/ai-image/adapters/": "adapter edits are the lifecycle contract's surface"
 ---
 
 # Provider Lifecycle Discipline