@event4u/agent-config 5.10.1 → 6.0.0

package/docs/contracts/profile-system.md

@@ -38,6 +38,7 @@ an ADR — the contract surface that ships in the wizard
 ```yaml
 profile:
   id: developer
+  packs: [engineering-base]              # base capability packs (scoped projection)
   audience:
     label: "IC engineer"
     readme_anchor: "developer"          # selects README first-screen block
@@ -50,6 +51,26 @@ profile:
     docs_first_pointer: "docs/getting-started-by-role.md#developer"
 ```
 
+### `packs` — base capability packs (6.0.0-B)
+
+`profile.packs` is the **base capability-pack set** the projector resolves
+when `projection.mode: scoped` (ADR-040). At projection time the selected set
+is `profile.packs ∪ runtime.active_packs` (the session overlay), expanded over
+the `requires` graph ([`capability-packs.md`](capability-packs.md)); the
+overlay can only **widen**, so `packs` is the floor of the default surface.
+
+**`skills_hint` is a guarantee, not a teaser.** Every skill named in a
+profile's `skills_hint` MUST resolve from that profile's `packs` (the
+self-sufficient contract). A profile must not advertise a skill its base packs
+hide. The six seed profiles satisfy this; the budget lint in 6.0.0-C will
+enforce it in CI.
+
+> Resolved by AI council (claude-sonnet-4-5 + gpt-4o, 2026-06-03): Option A
+> (self-sufficient base packs) for 6.0.0-B — no new mechanism, opt-in, default
+> stays `legacy-all`. Deferred to 6.0.0-C: a `skills_discoverable` field,
+> reactive just-in-time pack activation, the per-pack budget lint, and
+> telemetry.
+
 Per [ADR-010](../decisions/ADR-010-profile-pack-preset-boundary.md), a
 profile **MAY** set `defaults.preset_id` but **MAY NOT** set any
 preset-owned knob directly. The lint task (`task lint-config-schema`)
@@ -85,6 +106,7 @@ After the loader runs, the session has:
 ```python
 {
   "id": "developer",
+  "packs": ["engineering-base"],
   "audience": {"label": "IC engineer", "readme_anchor": "developer"},
   "preset_id": "balanced",
   "personas": ["reviewer", "security"],
@@ -127,7 +149,10 @@ Only changes to the **seed set** require an ADR.
 
 - This contract does **not** define preset knobs. See
   [`config-presets.md`](config-presets.md).
-- It does **not** define packs. See `workflow-packs.md` (Phase 2 item 7).
+- It declares **base capability packs** (`profile.packs`, 6.0.0-B) for scoped
+  projection, but does **not** define workflow-pack *bundles*. See
+  [`capability-packs.md`](capability-packs.md) (capability layer) and
+  `workflow-packs.md` (bundle layer).
 - It does **not** override `rule_loading_tier`. The rule-tier loader keeps
   its independent axis per
   [`cost-profile-defaults.md`](cost-profile-defaults.md).

package/docs/contracts/provider-lifecycle.md

@@ -6,7 +6,7 @@ keep-beta-until: 2026-08-15
 # Provider Lifecycle — AI Video / Image / Audio Adapter Tiers
 
 > **Status:** active · **Stability:** beta · **Owner:** universal-platform-refinement Phase 3
-> · **Linter:** none (declarative contract; enforced agent-in-the-loop via [`provider-lifecycle-discipline`](../../.agent-src/rules/provider-lifecycle-discipline.md))
+> · **Linter:** none (declarative contract; enforced agent-in-the-loop via [`provider-lifecycle-discipline`](../../dist/agent-src/rules/provider-lifecycle-discipline.md))
 > · **Source-of-truth template:** [`agents/templates/.ai-video.xml.example`](../../agents/templates/.ai-video.xml.example)
 
 Locks the lifecycle tagging used on every adapter under
@@ -116,7 +116,7 @@ snapshot — older snapshots live in git history, not in this file.
 
 ## See also
 
-- [`provider-lifecycle-discipline`](../../.agent-src/rules/provider-lifecycle-discipline.md) — the tier-2 routing rule that surfaces this contract when a provider is touched.
-- [`scripts/ai-video/lib/adapter-contract.md`](../../scripts/ai-video/lib/adapter-contract.md) — the four-method shell contract every adapter implements.
+- [`provider-lifecycle-discipline`](../../dist/agent-src/rules/provider-lifecycle-discipline.md) — the tier-2 routing rule that surfaces this contract when a provider is touched.
+- [`scripts/ai-video/lib/adapter-contract.md`](../../src/scripts/ai-video/lib/adapter-contract.md) — the four-method shell contract every adapter implements.
 - [`agents/templates/.ai-video.xml.example`](../../agents/templates/.ai-video.xml.example) — operator-facing provider configuration template (carries the tier tag inline).
 - [`agents/settings/policies/media/README.md`](../../agents/settings/policies/media/README.md) — the agent-in-the-loop enforcement model this contract participates in.

package/docs/contracts/release-pr-gating.md

@@ -1,3 +1,8 @@
+---
+stability: beta
+keep-beta-until: 2026-09-04
+---
+
 # Release-PR Gating Contract
 
 > **Status:** active · **Owner:** maintainer (`scripts/release.py`) · **Opened:** 2026-05-26

package/docs/contracts/release-trunk-sync.md

@@ -20,7 +20,7 @@ Every tagged release (`X.Y.Z`) **fast-forwards `main` to the tag's
 commit as the final step of the release pipeline**. No exceptions. No
 grace period.
 
-The fast-forward is owned by [`scripts/release.py`](../../scripts/release.py)
+The fast-forward is owned by [`scripts/release.py`](../../src/scripts/release.py)
 and runs after the GitHub Release is published. The release pipeline
 is **not green** until `main == <new-tag>` at the remote.
 
@@ -55,7 +55,7 @@ the release.
 
 ## CI Gate (P1.3)
 
-[`scripts/check_release_trunk_sync.py`](../../scripts/check_release_trunk_sync.py)
+[`scripts/check_release_trunk_sync.py`](../../src/scripts/check_release_trunk_sync.py)
 runs on every `release/X.Y.Z` branch (detected by `git rev-parse
 --abbrev-ref HEAD` matching `^release/\d+\.\d+\.\d+$`).
 
@@ -97,7 +97,7 @@ not orphan the convention.
 
 ## See also
 
-- [`scripts/release.py`](../../scripts/release.py) — release pipeline owner.
+- [`scripts/release.py`](../../src/scripts/release.py) — release pipeline owner.
 - [`.github/workflows/release-guard.yml`](../../.github/workflows/release-guard.yml)
   — tag/version-file integrity gate (orthogonal: this contract handles
   trunk position, release-guard handles version-string integrity).

package/docs/contracts/review-lens-schema.md

@@ -0,0 +1,59 @@
+# Review-lens routing schema
+
+Defines what a review **lens** is, the metadata it carries, and which review
+skills are **entry skills** (invoked directly / by task) vs **dispatched lenses**
+(run only under an orchestrator). This is the contract the Phase-3 consolidation
+scan validates merge candidates against — **proposal/schema only; no skill is
+merged here.**
+
+## Lens vs entry skill
+
+```
+A LENS IS DISPATCHED, NEVER A FRONT DOOR.
+AN ENTRY SKILL IS INVOKED; A LENS IS ORCHESTRATED.
+DO NOT MERGE ACROSS THE LINE.
+```
+
+- **Entry review skill** — a reviewer the user (or a task) invokes directly. It
+  produces a standalone verdict and may itself orchestrate. Examples:
+  `code-review` (also the `/review-changes` consolidator), `adversarial-review`,
+  `authz-review`, `design-review`, `privacy-review`, `receiving-code-review`,
+  `requesting-code-review`, `skill-reviewer`.
+- **Dispatched lens** — a specialized reviewer that runs **only** under an
+  orchestrator (`/review-changes` dispatches five, each on the same diff +
+  context, independently, then consolidates). Examples: `judge-bug-hunter`,
+  `judge-security-auditor`, `judge-test-coverage`, `judge-code-quality`,
+  `architecture-review-lens`.
+- **Router** — picks reviewers by path + risk, dispatches nothing itself:
+  `review-routing`.
+
+## Schema — the metadata a lens carries
+
+| Field | Meaning |
+|---|---|
+| `id` | the lens skill slug (`judge-bug-hunter`, `architecture-review-lens`, …) |
+| `scope` | the single dimension it reviews (correctness · security · test-coverage · quality · architecture) |
+| `constraints` | what it MUST flag and what it must NOT (stay in its lane — a bug lens does not opine on style) |
+| `dispatched_by` | the orchestrator that runs it (`/review-changes`; `/judge` for solo/steps) |
+| `runs` | `independently` — same diff + task context, no cross-talk between lenses |
+| `consolidated_by` | who merges the verdicts (`code-review` for `/review-changes`) |
+| `verdict_shape` | findings list with severity; no merge/ship decision (the human decides) |
+
+## The dispatch contract (what Phase 3 validates against)
+
+1. A **lens never becomes a front door** — it has no standalone command and is
+   not in a profile's `commands_hint`. Merging a lens into an entry skill (or
+   promoting it to a command) breaks the orchestrated-review model.
+2. A **lens stays single-scope** — consolidation must not fuse two lenses into
+   one multi-scope reviewer (that re-creates the monolith `/review-changes` was
+   built to decompose).
+3. An **entry skill may absorb another entry skill** only when both are
+   user-invoked and their scopes genuinely overlap — but never absorb a lens.
+4. The **consolidator** (`code-review`) is load-bearing — it is the entry skill
+   that both stands alone and merges lens verdicts; it is not a merge candidate.
+
+## See also
+
+- [`/review-changes`](../../src/domains/engineering-base/review-changes/command.md) — the five-judge dispatcher.
+- [`skill-family-map.yml`](skill-family-map.yml) — `family: review-judging` carries these skills; Phase 3 reads `candidate_*` against this contract.
+- [`governance.md`](../governance.md) — the lifecycle + Single-Source-of-Truth rules these skills age under.

package/docs/contracts/role-experience.md

@@ -96,11 +96,14 @@ Every `id` must resolve to an existing skill under `packages/<pack>/.agent-src.u
 
 ## Versioning + status
 
-- `status: draft` — scaffold only, no recruit-session evidence yet. Launcher hides drafts in the default view.
-- `status: beta` — scaffold + at least one recruit session backs the first-task choice. Launcher surfaces with a `beta` badge.
-- `status: stable` — at least two recruit sessions concurred on the role's first tasks and the friction-inventory landed at ≤ 3 P0 items.
+- `status: draft` — scaffold only, no validation evidence yet. Launcher hides drafts in the default view.
+- `status: beta-internal` — scaffold + maintainer internal-authoring / dogfooding basis; **no external recruit session yet**. `recruit_session_ref` stays `null`. Launcher surfaces with an `internal beta` badge. This tier exists so a role can be *used* without claiming external validation it does not have. The basis is recorded in [`agents/roles/EVIDENCE_BASIS.md`](../../agents/roles/EVIDENCE_BASIS.md). It is the honest ceiling for a role that has never been in front of an external user.
+- `status: beta` — scaffold + at least one recruit session backs the first-task choice. Requires a non-null `recruit_session_ref`. Launcher surfaces with a `beta` badge.
+- `status: stable` — at least two recruit sessions concurred on the role's first tasks and the friction-inventory landed at ≤ 3 P0 items. Requires a non-null `recruit_session_ref`.
 
-Status promotion is a maintainer decision, captured in the `recruit_session_ref` frontmatter trail. Demotion (stable → beta) happens when a follow-up recruit session contradicts the existing first-task choice.
+Status promotion is a maintainer decision. `draft → beta-internal` is an internal-authoring promotion (no recruit session needed); `beta-internal → beta → stable` is the **external-validation** ladder and is gated on the `recruit_session_ref` trail. Demotion (stable → beta) happens when a follow-up recruit session contradicts the existing first-task choice.
+
+> **Why two beta tiers (AI-council convergence, claude-sonnet-4-5 + gpt-4o, 2026-06-08, design mode):** the recruit sessions test a boundary the self-improvement loop is blind to — *can a cold-start external user get in at all?* — so they are kept as an optional future activity, never faked and never cancelled. `beta-internal` lets downstream work proceed and the launcher surface the roles **today**, honestly labelled, while `beta`/`stable` stay reserved for real external evidence. Fabricating a `recruit_session_ref` to reach `beta` is the one move this contract forbids.
 
 ## Lint pass (deferred to Phase 3 Step 6)
 
@@ -110,9 +113,10 @@ The lint pass (`task lint-role-experiences`, wired into `task ci`) asserts:
 2. Every `prompts/<name>.md` has the four required frontmatter keys (`name`, `intent`, `inputs`, `output_shape`) plus `skill_hint`.
 3. Every `skills.yml#skills[].id` resolves to an existing skill.
 4. Every `recommended_packs[]` entry resolves to an existing pack manifest.
-5. Status promotion never happens without a non-null `recruit_session_ref`.
+5. `status` is one of `draft | beta-internal | beta | stable`.
+6. `status: beta` and `status: stable` require a **non-null** `recruit_session_ref` (external-validation gate). `draft` and `beta-internal` may keep `recruit_session_ref: null`.
 
-Lint pass code lives at `scripts/lint_role_experiences.py`, ≤ 200 LOC. Until the lint pass ships, the scaffolds in Phase 3 Steps 2–4 are hand-validated against this contract.
+Lint pass code lives at `scripts/lint_role_experiences.py`, ≤ 200 LOC.
 
 ## Open questions (Phase 3 optional council pass)
 

package/docs/contracts/router-blending.md

@@ -6,7 +6,7 @@ keep-beta-until: 2026-08-13
 
 # Router Blending — Cross-Domain Skill Mix Contract
 
-> **Status:** beta · **Owner:** router governance · **Depends on:** [`universal-skills.md`](universal-skills.md), the user-types axis (axis seeds — schema field defined under [`skill.schema.json`](../../scripts/schemas/skill.schema.json) → `recommended_for_user_types`).
+> **Status:** beta · **Owner:** router governance · **Depends on:** [`universal-skills.md`](universal-skills.md), the user-types axis (axis seeds — schema field defined under [`skill.schema.json`](../../src/scripts/schemas/skill.schema.json) → `recommended_for_user_types`).
 
 ## Purpose
 
@@ -67,5 +67,5 @@ Wired into `task ci` as a non-blocking warning until the user-types axis stabili
 ## Related
 
 - [`universal-skills.md`](universal-skills.md) — the always-loaded floor under the blend.
-- [`skill.schema.json`](../../scripts/schemas/skill.schema.json) — `recommended_for_user_types` field that carries the tags.
+- [`skill.schema.json`](../../src/scripts/schemas/skill.schema.json) — `recommended_for_user_types` field that carries the tags.
 - [`tests/eval/corpus-non-dev.yaml`](../../tests/eval/corpus-non-dev.yaml) — empirical validation of the blends.

package/docs/contracts/rule-classification.md

@@ -76,7 +76,7 @@ in place per P2.2 playbook. Loaded by router on trigger match.
 |---|---:|---|---|
 | architecture | 2491 | 3 | Behavioural; project-organization gates |
 | artifact-drafting-protocol | 2957 | 2a | Iron-Law-shaped Understand→Research→Draft sequence |
-| augment-source-of-truth | 2439 | 1 | Behavioural Iron Law: never edit generated dirs |
+| source-of-truth | 2439 | 1 | Behavioural Iron Law: never edit generated dirs |
 | autonomous-execution | 5631 | 3 | Trivial-vs-blocking decision; tier-1 on `balanced` |
 | context-hygiene | 3811 | 1 | Behavioural Iron Law: 3-failure stop, fresh-chat trigger |
 | downstream-changes | 2940 | 2b | Behavioural: every code edit |

package/docs/contracts/rule-router.md

@@ -93,9 +93,9 @@ start. Deterministic key order, sorted lists, stable across runs.
              "direct-answers", "language-and-tone", "no-cheap-questions",
              "non-destructive-by-default", "scope-control",
              "verify-before-complete"],
-  "tier_1": [{"id": "augment-source-of-truth",
+  "tier_1": [{"id": "source-of-truth",
               "triggers": [{"path_prefix": "agents/"},
-                           {"path_prefix": ".agent-src/"}],
+                           {"path_prefix": "dist/agent-src/"}],
               "routes_to": ["skill:agent-docs-writing"]}],
   "tier_2": [/* same shape as tier_1 */],
   "profiles": {
@@ -141,7 +141,7 @@ lean_projection:
 
 Revert procedure (one flip, no code change): set `lean_projection.mode:
 eager-all`, run `task generate-tools` (regenerates `.claude/`, `.cursor/`,
-`.clinerules/`, `.windsurfrules`) + `task sync` (`.agent-src/`, `.augment/`).
+`.clinerules/`, `.windsurfrules`) + `task sync` (`dist/agent-src/`, `.augment/`).
 The thin projector (Phase 3.1) MUST honour this key; with it absent or
 `eager-all` the projector behaves exactly as today. Default stays
 `eager-all` so the migration is opt-in and reversible by one line.

package/docs/contracts/safety-model.md

@@ -75,7 +75,7 @@ A profile may only reference an id from the table below.
 | Policy | Behaviour | Floor interaction |
 |---|---|---|
 | `deny` | The agent refuses. Numbered-option block surfaces the refusal and the rationale field; no override path. | `deny` is the default for every floor domain — it cannot be relaxed. |
-| `ask` | The agent stops and asks a single numbered question per [`user-interaction`](../../.agent-src/rules/user-interaction.md). One question per turn. | `ask` is the default for every floor domain in a profile that has not opted out — the floor remains operative even when `policy=allow` is set elsewhere. |
+| `ask` | The agent stops and asks a single numbered question per [`user-interaction`](../../dist/agent-src/rules/user-interaction.md). One question per turn. | `ask` is the default for every floor domain in a profile that has not opted out — the floor remains operative even when `policy=allow` is set elsewhere. |
 | `allow` | The agent proceeds without asking. Trivial-question suppression applies. | `allow` is **forbidden** on any domain whose `Floor reference` column is non-empty. Linter rejects it. |
 
 The legacy single switch (`personal.autonomy`) is preserved as a
@@ -91,7 +91,7 @@ Order (last writer wins, subject to the Iron Floor):
 3. Active pack's profile (if `--pack <id>` is active).
 4. `.agent-settings.yml` user override under `profile.safety.domains`.
 
-The explain command at [`explain config`](../../.agent-src/scripts/agent-config)
+The explain command at [`explain config`](../../dist/agent-src/scripts/agent-config)
 (Phase 1 Item 3 deliverable) surfaces the resolved policy per domain,
 with the writer source per row.
 

package/docs/contracts/session-profile-overlay.md

@@ -42,9 +42,9 @@ runtime:
 
 `/profile activate <name…>` accepts, for each name:
 
-1. a **session-profile alias** from `config/discovery/session-profiles.yml`
+1. a **session-profile alias** from `src/config/discovery/session-profiles.yml`
    (`developer`, `po`, `finance`, `gtm`, `content`) → its seed pack list, or
-2. a **raw pack id** from `config/discovery/packs.yml` (`laravel`, `php`, …).
+2. a **raw pack id** from `src/config/discovery/packs.yml` (`laravel`, `php`, …).
 
 Multiple names union their closures. The seed set's transitive
 `requires_hint` closure is expanded before writing. **Only installed packs
@@ -83,6 +83,14 @@ one-line "from inactive pack X" notice. Hard execution-gating is deferred
   concurrent reader never sees a half-written overlay.
 - **Closure self-heal** — a closure dependency that is not installed is
   dropped from the written set with a note, never blocking activation.
+- **Set-only — precedence intentionally undefined.** The overlay is an
+  order-independent **union of pack ids**: no precedence, no scalar
+  "audience hint", no ordering. `set_overlay` always writes `sorted(set(...))`,
+  and the static definitions are frozen at the data layer by
+  `scripts/lint_profile_overlay_set_only.py` (aliases resolve only to pack-id
+  sets; no profile/pack file declares a scalar `active_packs` or a
+  `precedence`/`priority`/`order` key). A future scalar-precedence regression
+  fails the build rather than silently re-introducing a precedence concept.
 
 ## Reconciliation with the existing `--profile=<id>` install flag
 
@@ -108,10 +116,39 @@ axis (ADR-010 addendum) so it adds no new axis despite sharing the word.
 
 - Library + CLI: `scripts/config/session_profiles.py`
   (`activate` · `deactivate` · `show` · `surface` · `stale-notice`).
-- Aliases: `config/discovery/session-profiles.yml`.
+- Aliases: `src/config/discovery/session-profiles.yml`.
 - Hook: `scripts/profile_staleness_hook.py` (session_start staleness notice).
 - Tests: `tests/test_session_profiles.py`.
 
+## Plain status surface
+
+`/profile show --plain` renders the active-overlay state in plain language for a
+**non-technical employee** — "which profile, what it surfaces/hides, why the
+agent behaves differently" — over the existing `show` state, with **zero new
+overlay logic**.
+
+It is a **deterministic, template-based** render (`session_profiles.format_plain_status`),
+**never LLM-generated**: a pure function of the `show` JSON
+(`active_packs`, `commands_shown`, `skills_shown`, `hidden_total`). This is a hard
+constraint — it removes any hidden-pack-name leak or hallucination surface, and
+lets the golden tests pin the output byte-for-byte.
+
+The render says, for an active overlay:
+
+1. the active profile/pack name(s);
+2. how many commands + skills you'll **see**;
+3. how many items are **hidden** behind packs you haven't turned on (this doubles
+   as the "what changed vs the full surface" line);
+4. that the overlay **persists across sessions** until `/profile deactivate`.
+
+With no overlay it states the full surface is active.
+
+**Staleness is rendered as persistence, not an age-in-days.** The overlay
+(`runtime.active_packs`) carries no timestamp, and adding one would change overlay
+semantics — out of scope for this surface. A future age-in-days render is a
+separate, semantics-touching change (an overlay set-time), not a plain-render
+concern. The session-boundary reminder remains the `stale-notice` hook's job.
+
 ## See also
 
 - [`ADR-010 addendum`](../decisions/ADR-010-profile-pack-preset-boundary.md) — overlay ≠ fifth axis.

package/docs/contracts/settings-api.md

@@ -53,7 +53,7 @@ token for optimistic locking. Merge order (per **ADR-020** + roadmap
 `road-to-global-only-install` § Phase 2.2):
 
 ```text
-defaults (config/agent-settings.template.yml)
+defaults (src/config/agent-settings.template.yml)
   < global (~/.event4u/agent-config/settings/.agent-settings.yml)
     < project (<projectRoot>/settings/.agent-settings.yml, optional)
 ```
@@ -132,7 +132,7 @@ when the file is missing (wizard pre-fill case).
 
 ### `GET /api/v1/user-md/template`
 
-Returns the package-shipped template (`templates/agent-user.md`) when
+Returns the package-shipped template (`src/templates/agent-user.md`) when
 present, else 204 No Content. Wizard uses this to pre-fill the textarea
 on first run.
 
@@ -205,4 +205,4 @@ Round-trip fixtures live in `tests/server/fixtures/`. Every endpoint
 above has a happy-path test in `tests/server/<endpoint>.test.ts` and at
 least one error-path test (409 or 422). The schema↔template parity
 test (`tests/server/schemas/parity.test.ts`) is the gate that keeps
-`settingsSchema` and `config/agent-settings.template.yml` in lockstep.
+`settingsSchema` and `src/config/agent-settings.template.yml` in lockstep.

package/docs/contracts/settings-sync-yaml-subset.md

@@ -6,9 +6,9 @@ keep-beta-until: 2026-08-12
 # Settings-sync YAML subset
 
 **Purpose.** Pin the YAML feature set that `.agent-settings.yml` and
-`config/agent-settings.template.yml` may use, so contributors can cite a
+`src/config/agent-settings.template.yml` may use, so contributors can cite a
 contract instead of inferring it from
-[`scripts/sync_yaml_rt.py`](../../scripts/sync_yaml_rt.py) source. The
+[`scripts/sync_yaml_rt.py`](../../src/scripts/sync_yaml_rt.py) source. The
 sync engine ([ADR](adr-settings-sync-engine.md)) is a custom stdlib-only
 round-trip parser/emitter; staying inside the subset below is what
 keeps user-line preservation (every byte of every user line round-trips
@@ -135,5 +135,5 @@ document is updated in the same commit.
 
 - [`docs/contracts/adr-settings-sync-engine.md`](adr-settings-sync-engine.md) — decision record for the stdlib-only engine.
 - [`docs/guidelines/agent-infra/layered-settings.md`](../guidelines/agent-infra/layered-settings.md) § Sync rules — the additive-merge-with-user-line-preservation contract this subset implements.
-- [`scripts/sync_yaml_rt.py`](../../scripts/sync_yaml_rt.py) — implementation; module docstring mirrors this file.
-- [`scripts/sync_agent_settings.py`](../../scripts/sync_agent_settings.py) — CLI driver and exit-code contract.
+- [`scripts/sync_yaml_rt.py`](../../src/scripts/sync_yaml_rt.py) — implementation; module docstring mirrors this file.
+- [`scripts/sync_agent_settings.py`](../../src/scripts/sync_agent_settings.py) — CLI driver and exit-code contract.

package/docs/contracts/skill-bundled-assets.md

@@ -0,0 +1,62 @@
+---
+stability: beta
+keep-beta-until: 2026-09-01
+---
+
+# Skill-bundled assets — `skills/<name>/{scripts,data}/` delivery contract
+
+> Sanctions and pins the pattern introduced by ADR-061: a skill may bundle
+> executable Python under `scripts/` and corpus/config files under `data/`
+> (plus `references/` docs), and those assets **reach and run at consumer
+> runtime**. Locked by `tests/test_skill_bundled_assets.py`.
+
+## The delivery chain (verified)
+
+| Stage | Mechanism | Guarantee |
+|---|---|---|
+| 1. Authoring | `src/skills/<name>/{scripts,data,references}/` | Source of truth; `.md` prose is condensed, everything else is data. |
+| 2. Sync | `src/scripts/condense.py::sync_non_md` (`task sync`) | Every non-`.md` skill asset is copied byte-identical into `dist/agent-src/skills/<name>/…`. |
+| 3. Packaging | `package.json` ships `dist/agent-src/` | Assets are inside the published npm artifact. |
+| 4. Install | `install.py::_copy_dir_dereferencing_symlinks` | Whole skill directories (including `scripts/` + `data/`) are deployed to the consumer scope, e.g. `~/.claude/skills/<name>/…`, with symlinks dereferenced so the copy survives npx cache eviction. |
+
+## Invocation contract — skill-relative, never project-relative
+
+```
+A BUNDLED SCRIPT RESOLVES EVERY SIBLING ASSET VIA ITS OWN FILE LOCATION
+(Path(__file__).resolve().parent…), NEVER VIA $PWD OR THE PROJECT ROOT.
+```
+
+- The agent invokes bundled scripts with an **absolute or skills-root-
+  relative path**: `python3 <skills-root>/<name>/scripts/<tool>.py …`.
+  `<skills-root>` = `~/.claude/skills` (Claude Code consumer install),
+  the tool-specific projection dir for other agents, or `src/skills`
+  inside this repo.
+- The consumer's working directory is **arbitrary** — scripts MUST work
+  from any cwd. Writing output uses caller-supplied paths (flags), never
+  cwd-implicit locations, except where a flag explicitly defaults to cwd.
+- Cross-skill references (one skill invoking another's bundled engine)
+  resolve via the common skills root (e.g. `design-intelligence` passing
+  its `data/manifest.json` to `corpus-grounding/scripts/ground.py`).
+  Sibling deployment is guaranteed by stage 4 (whole-directory copies of
+  every installed skill into one root).
+- Bundled scripts are **pure stdlib** unless the skill's frontmatter
+  declares an `execution` block gating anything heavier
+  (`runtime-safety`); no network and no subprocess by default.
+
+## What this contract does NOT cover
+
+- Repo-root `scripts/` (e.g. `council_cli.py`) — **maintainer-only**,
+  never shipped to consumers. Do not confuse the two layers.
+- `.md` files inside skills — they pass through condensation (prose is
+  rewritten); do not put load-bearing machine-read data in skill `.md`.
+- Binary assets — allowed by the chain, but keep corpora text-based
+  (CSV/JSON) so diffs and license review stay possible.
+
+## Regression lock
+
+`tests/test_skill_bundled_assets.py` builds a fixture skill
+(`scripts/probe.py` + `data/fixture.csv`), simulates the install copy into
+a temp skills root, invokes the script from an unrelated cwd, and asserts
+it finds its sibling data — plus runs the real `corpus-grounding` engine
+from a foreign cwd. If that test goes red, the grounding layer is theater
+at consumer runtime (road-to-frontend-design-intelligence Step 1.0).

package/docs/contracts/skill-discovery.md

@@ -19,7 +19,7 @@ new always-loaded layer (that would fail the Phase-1 leaner-core premise).
 
 | Signal | Source | Used for |
 |---|---|---|
-| Skill catalog | `.agent-src/skills/*/SKILL.md` frontmatter (`name`, `description`, `domain`) | candidate universe + `domain` category |
+| Skill catalog | `dist/agent-src/skills/*/SKILL.md` frontmatter (`name`, `description`, `domain`) | candidate universe + `domain` category |
 | Role shortlist | `agents/roles/<role>/skills.yml` (priority-ordered `id` + `why`) | `most-useful-for-role` |
 | Local analytics | `~/.event4u/agent-config/workspace/analytics/events.jsonl` (`event`, `data.role`, `data.task`, optional `data.skill`) | `recently-adopted`, `popular-in-role` |
 

package/docs/contracts/skill-distribution-channels.md

@@ -1,8 +1,13 @@
+---
+stability: beta
+keep-beta-until: 2026-09-04
+---
+
 # Skill Distribution Channels — canonical per AI tool
 
 **Status:** Active (locked 2026-05-25 via Phase A of `road-to-clean-skill-distribution-channels.md`)
 **Owner:** maintainer-team
-**Inputs:** [`agents/evidence/audits/2026-05-distribution-channels/`](../../agents/evidence/audits/2026-05-distribution-channels/) (audits 01 + 02), AI Council convergence `agents/runtime/council/responses/2026-05-25-canonical-channel.json` (claude-sonnet-4-5 + gpt-4o, 2026-05-25)
+**Inputs:** [`agents/evidence/audits/2026-05-distribution-channels/`](../../agents/evidence/audits/2026-05-distribution-channels/) (audits 01 + 02), AI Council convergence on the canonical-channel question (claude-sonnet-4-5 + gpt-4o, 2026-05-25 — session artefact auto-pruned; the decision body is inlined in this contract)
 
 ## Rule
 

package/docs/contracts/skill-dry-run.md

@@ -29,7 +29,7 @@ IT IS NOT A GUARANTEE OF SIDE-EFFECT-FREENESS FOR SKILLS WITH AN `execution` BLO
 Preview reads declared intent — it does not run the skill, does not dry-run its commands, and cannot
 prove a skill is harmless. It tells you what the skill *says* it will touch, so you can decide whether
 to run it. For `execution: manual` skills (the default), it states plainly: **instructional only — no
-automatic execution** (per [`runtime-safety`](../../.agent-src/rules/runtime-safety.md): `manual` is
+automatic execution** (per [`runtime-safety`](../../dist/agent-src/rules/runtime-safety.md): `manual` is
 instructional, `assisted` must propose before executing).
 
 ## Surface
@@ -43,5 +43,5 @@ SKILL.md degrades to a **structured error**, never a crash.
 
 ## Implementation
 
-`scripts/skill_preview.py` (≤ 250 LOC). Read-only over `.agent-src/skills/<name>/SKILL.md`. No network,
+`scripts/skill_preview.py` (≤ 250 LOC). Read-only over `dist/agent-src/skills/<name>/SKILL.md`. No network,
 no execution. Coverage: `tests/test_skill_preview.py`.