@event4u/agent-config 3.3.0 → 4.2.0

package/docs/contracts/multi-tool-projection-fidelity.md

@@ -8,11 +8,11 @@ keep-beta-until: 2026-08-13
 **Status:** beta · **Phase 4 of the `step-1-v2-feedback-followup`
 roadmap** (under `agents/roadmaps/`).
 
-Names the **per-tool guarantees** the projection pipeline (`scripts/compress.py --sync` + `scripts/compress.py --generate-tools`) actually delivers. Byte-equivalence is not behaviour-fidelity — each consumer tool has its own frontmatter grammar, its own activation model, and its own surface for skills / rules / commands.
+Names the **per-tool guarantees** the projection pipeline (`scripts/condense.py --sync` + `scripts/condense.py --generate-tools`) actually delivers. Byte-equivalence is not behaviour-fidelity — each consumer tool has its own frontmatter grammar, its own activation model, and its own surface for skills / rules / commands.
 
 ## Source of truth
 
-Every projection starts from `.agent-src/` (compressed) which is generated from `.agent-src.uncompressed/`. The projection layer **never** writes to source; it only reads.
+Every projection starts from `.agent-src/` (condensed) which is generated from `.agent-src.uncondensed/`. The projection layer **never** writes to source; it only reads.
 
 ## Per-tool projection map
 
@@ -96,7 +96,7 @@ These are **architectural facts**, not regressions. They are documented so insta
 2. **Windsurf single-file (`.windsurfrules`) strips per-rule frontmatter.** Legacy compatibility surface. The new `.windsurf/rules/*.md` per-rule files preserve the full frontmatter — consumers should prefer those.
 3. **Skills do not project to Cursor / Windsurf / Cline / Gemini / Copilot.** These tools have no native skill loader. Skill content reaches consumers indirectly via rule bodies and the `AGENTS.md` catalogue.
 4. **Augment historically did not load symlinked rules.** Default is to **copy** rules into `.augment/rules/`. Opt into symlinks via `augment.rules_use_symlinks: true` in `.agent-settings.yml`.
-5. **`task generate-tools` does not refresh `.augment/rules/`.** Only `task sync` (== `scripts/compress.py --sync`) copies rules into the Augment tree. Investigators who edit a rule, run only `generate-tools`, and then `ls .augment/rules/` will see stale state.
+5. **`task generate-tools` does not refresh `.augment/rules/`.** Only `task sync` (== `scripts/condense.py --sync`) copies rules into the Augment tree. Investigators who edit a rule, run only `generate-tools`, and then `ls .augment/rules/` will see stale state.
 
 ## Acceptance criteria for this contract
 
@@ -108,7 +108,7 @@ These are **architectural facts**, not regressions. They are documented so insta
 
 ## Related
 
-- [`source-projection`](../architecture/source-projection.md) — pipeline A (source compression)
+- [`source-projection`](../architecture/source-projection.md) — pipeline A (source condensation)
 - [`augment-projection`](../architecture/augment-projection.md) — pipeline B (Augment-specific)
 - [`multi-tool-projection`](../architecture/multi-tool-projection.md) — pipeline C (the per-tool emitters)
 - [`rule-router`](rule-router.md) — the `triggers:` / `routes_to:` grammar this contract pins

package/docs/contracts/namespace.md

@@ -59,10 +59,10 @@ skipped by the linter.
 
 | Type | Source path | Naming nuance |
 |---|---|---|
-| Skill | `.agent-src.uncompressed/skills/<name>/SKILL.md` | Directory name == frontmatter `name`. |
-| Rule | `.agent-src.uncompressed/rules/<name>.md` | Filename stem == frontmatter `id` (when present). |
-| Command | `.agent-src.uncompressed/commands/<name>.md` or `<group>/<verb>.md` | Slash-command invocation `<name>` or `<group>:<verb>`. |
-| Persona | `.agent-src.uncompressed/personas/<name>.md` | Cited from skill frontmatter `personas:` list. |
+| Skill | `.agent-src.uncondensed/skills/<name>/SKILL.md` | Directory name == frontmatter `name`. |
+| Rule | `.agent-src.uncondensed/rules/<name>.md` | Filename stem == frontmatter `id` (when present). |
+| Command | `.agent-src.uncondensed/commands/<name>.md` or `<group>/<verb>.md` | Slash-command invocation `<name>` or `<group>:<verb>`. |
+| Persona | `.agent-src.uncondensed/personas/<name>.md` | Cited from skill frontmatter `personas:` list. |
 
 Sub-namespacing (`commit/in-chunks.md` →  `/commit:in-chunks`) uses
 the same charset rules per segment; the joining colon is implicit.

package/docs/contracts/orchestration-dsl-v1.md

@@ -13,7 +13,7 @@ that re-runs the same step sequence with the same inputs.
 **Scope.** Defines the file location, top-level shape, step kinds,
 input / output wiring, and the linter contract. Does **not** define
 the runtime semantics of each step kind — those live in the
-[`/orchestrate`](../../.agent-src.uncompressed/commands/orchestrate.md)
+[`/orchestrate`](../../.agent-src.uncondensed/commands/orchestrate.md)
 command and the `work_engine` directive modules it delegates to.
 
 Last refreshed: 2026-05-11.
@@ -95,10 +95,10 @@ outputs:
 
 | `kind` | `ref` resolves to | Runtime |
 |---|---|---|
-| `skill` | `.agent-src.uncompressed/skills/<ref>/SKILL.md` | Dispatched via `work_engine` directive matching the skill's domain. |
-| `command` | `.agent-src.uncompressed/commands/<ref>.md` | Same dispatch path the slash-command takes when typed by the user. |
-| `persona` | `.agent-src.uncompressed/personas/<ref>.md` | Sets `roles.active_role` for the next dependent step; does not produce its own output. |
-| `subagent` | `subagent-orchestration` mode name | Spawned per [`subagent-orchestration`](../../.agent-src.uncompressed/skills/subagent-orchestration/SKILL.md). |
+| `skill` | `.agent-src.uncondensed/skills/<ref>/SKILL.md` | Dispatched via `work_engine` directive matching the skill's domain. |
+| `command` | `.agent-src.uncondensed/commands/<ref>.md` | Same dispatch path the slash-command takes when typed by the user. |
+| `persona` | `.agent-src.uncondensed/personas/<ref>.md` | Sets `roles.active_role` for the next dependent step; does not produce its own output. |
+| `subagent` | `subagent-orchestration` mode name | Spawned per [`subagent-orchestration`](../../.agent-src.uncondensed/skills/subagent-orchestration/SKILL.md). |
 
 ## Interpolation
 
@@ -147,7 +147,7 @@ majors.
 
 ## Cross-references
 
-- Command surface: [`/orchestrate`](../../.agent-src.uncompressed/commands/orchestrate.md).
+- Command surface: [`/orchestrate`](../../.agent-src.uncondensed/commands/orchestrate.md).
 - Linter: [`lint_orchestration_dsl.py`](../../scripts/lint_orchestration_dsl.py).
 - Runtime dispatcher precedent: [`implement-ticket-flow.md`](implement-ticket-flow.md).
-- Subagent runtime: [`subagent-orchestration`](../../.agent-src.uncompressed/skills/subagent-orchestration/SKILL.md).
+- Subagent runtime: [`subagent-orchestration`](../../.agent-src.uncondensed/skills/subagent-orchestration/SKILL.md).

package/docs/contracts/package-self-orientation.md

@@ -36,7 +36,7 @@ Four wings compose via [`cross-wing-handoff.md`](cross-wing-handoff.md)
 
 ## Tech stack
 
-Bash (install scripts) · Python 3.10+ (linters, compression, pytest) ·
+Bash (install scripts) · Python 3.10+ (linters, condensation, pytest) ·
 Markdown (all content) · Taskfile (`task ci/sync/test`) · GitHub
 Actions (`.github/workflows/`). Non-text inputs (PDF, DOCX, XLSX,
 images, audio) route through the
@@ -49,20 +49,20 @@ mechanics deep-dive: [`agents-md-tech-stack.md`](agents-md-tech-stack.md) (beta)
 
 | Directory | Purpose | Editable? |
 |---|---|---|
-| `.agent-src.uncompressed/` | Authoring layer — full verbose content | ✅ Yes — edit here |
-| `.agent-src/` | Compressed output — shipped in the package, consumed by agents | ❌ No — regenerated |
+| `.agent-src.uncondensed/` | Authoring layer — full verbose content | ✅ Yes — edit here |
+| `.agent-src/` | Condensed output — shipped in the package, consumed by agents | ❌ No — regenerated |
 | `.augment/` | Local projection of `.agent-src/` for Augment Code (gitignored) | ❌ No — regenerated |
 | `.claude/`, `.cursor/`, `.clinerules/`, `.windsurfrules` | Tool-specific projections | ❌ No — regenerated |
 | `agents/` | Package's own roadmaps, contexts, sessions | ✅ Yes |
 
 **Never edit `.agent-src/` or `.augment/` directly.** Edit
-`.agent-src.uncompressed/` and run `task sync` (or `task ci`) to
-compress + regenerate the tool directories.
+`.agent-src.uncondensed/` and run `task sync` (or `task ci`) to
+condense + regenerate the tool directories.
 
 ## Repository layout
 
 ```
-.agent-src.uncompressed/      ← edit here
+.agent-src.uncondensed/      ← edit here
   skills/       (150 skills)
   rules/        (58 rules)
   commands/     (103 commands)
@@ -73,10 +73,10 @@ compress + regenerate the tool directories.
 docs/guidelines/            (47 guidelines — reference material, not packaged)
 docs/contracts/             (kernel-membership, rule-router, rule-classification, …)
 docs/decisions/             (ADRs — kernel overrides, scope decisions)
-.agent-src/                 ← compressed output shipped in the package
+.agent-src/                 ← condensed output shipped in the package
 .agent-src/router.json      ← compiled router manifest (consumed at runtime)
 .augment/                   ← local projection for Augment Code (gitignored)
-scripts/                    ← install.sh, install.py, compress.py, linters
+scripts/                    ← install.sh, install.py, condense.py, linters
 tests/                      ← pytest (324 tests) + test_install.sh
 agents/                     ← this package's own roadmaps / sessions / contexts
 .github/workflows/          ← CI
@@ -88,7 +88,7 @@ agents/                     ← this package's own roadmaps / sessions / context
 |---|---|
 | `.agent-src/` must stay project-agnostic — no project names, domains, stacks | [`augment-portability`](../../.agent-src/rules/augment-portability.md) |
 | Root AGENTS.md + copilot-instructions.md must stay project-agnostic too | [`augment-portability`](../../.agent-src/rules/augment-portability.md) |
-| Edit `.agent-src.uncompressed/`, never `.agent-src/` or `.augment/` | [`augment-source-of-truth`](../../.agent-src/rules/augment-source-of-truth.md) |
+| Edit `.agent-src.uncondensed/`, never `.agent-src/` or `.augment/` | [`augment-source-of-truth`](../../.agent-src/rules/augment-source-of-truth.md) |
 | Skills must declare frontmatter, be self-contained, pass the linter | [`skill-quality`](../../.agent-src/rules/skill-quality.md) |
 | Size budgets for skills, rules, commands | [`size-enforcement`](../../.agent-src/rules/size-enforcement.md) |
 | Keep `.agent-src/` / `agents/` cross-refs in sync on add/rename/delete | [`docs-sync`](../../.agent-src/rules/docs-sync.md) |
@@ -99,7 +99,7 @@ agents/                     ← this package's own roadmaps / sessions / context
 Default-off. `telemetry.artifact_engagement.enabled: true` in
 `.agent-settings.yml` enables local-only JSONL logging. Redaction
 floor + pipeline:
-[`artifact-engagement-flow.md`](../../.agent-src.uncompressed/contexts/contracts/artifact-engagement-flow.md)
+[`artifact-engagement-flow.md`](../../.agent-src.uncondensed/contexts/contracts/artifact-engagement-flow.md)
 (beta). Rule:
 [`artifact-engagement-recording`](../../.agent-src/rules/artifact-engagement-recording.md).
 
@@ -111,7 +111,7 @@ surfaces matches as numbered options with a "run as-is" escape;
 Rule: [`command-suggestion-policy`](../../.agent-src/rules/command-suggestion-policy.md).
 Eligibility + scoring + hardening:
 [`adr-command-suggestion.md`](adr-command-suggestion.md) and
-[`command-suggestion-flow.md`](../../.agent-src.uncompressed/contexts/contracts/command-suggestion-flow.md)
+[`command-suggestion-flow.md`](../../.agent-src.uncondensed/contexts/contracts/command-suggestion-flow.md)
 (beta).
 
 ## Multi-agent tool support
@@ -126,7 +126,7 @@ layout + cloud-bundle pipeline:
 
 ## Contributing
 
-1. Edit inside `.agent-src.uncompressed/` or `scripts/` or `tests/` —
+1. Edit inside `.agent-src.uncondensed/` or `scripts/` or `tests/` —
    never in `.agent-src/`, `.augment/`, `.claude/`, `.cursor/`, etc.
 2. Run `task ci` locally. It must exit 0.
 3. Commit in logical chunks with Conventional Commits.

package/docs/contracts/persona-schema.md

@@ -7,7 +7,7 @@ keep-beta-until: 2026-08-12
 
 > **Status:** active · **Stability:** beta · **Owner:** road-to-better-skills-and-profiles Block A
 > · **Linter:** `scripts/skill_linter.py § lint_persona`
-> · **Source-of-truth dir:** `.agent-src.uncompressed/personas/`
+> · **Source-of-truth dir:** `.agent-src.uncondensed/personas/`
 > · **Council verdict:** iter-1 A-OQ1 (c) — hybrid
 
 Locks the canonical persona shape: a uniform frontmatter across both
@@ -116,9 +116,9 @@ The linter (A2 work) enforces:
 - description ≤ 160 chars
 
 Specialist authors must use the template at
-`.agent-src.uncompressed/personas/_template-specialist/persona.md`.
+`.agent-src.uncondensed/personas/_template-specialist/persona.md`.
 Core authors continue to use
-`.agent-src.uncompressed/templates/persona.md` (unchanged).
+`.agent-src.uncondensed/templates/persona.md` (unchanged).
 
 ## § 6 — Persona inventory (v1 snapshot)
 
@@ -148,7 +148,7 @@ loosening is non-breaking.
 
 - [`skill-domains`](skill-domains.md) — sister taxonomy for skills
 - [`role-contracts`](../guidelines/agent-infra/role-contracts.md) — workflow-mode axis personas compose with
-- `.agent-src.uncompressed/personas/README.md` — authoring entry point + Core-6 cast
-- `.agent-src.uncompressed/personas/_template-specialist/persona.md` — specialist starter
-- `.agent-src.uncompressed/templates/persona.md` — core starter (v1, unchanged)
+- `.agent-src.uncondensed/personas/README.md` — authoring entry point + Core-6 cast
+- `.agent-src.uncondensed/personas/_template-specialist/persona.md` — specialist starter
+- `.agent-src.uncondensed/templates/persona.md` — core starter (v1, unchanged)
 - `road-to-better-skills-and-profiles.md` — Block A (schema lock + extension)

package/docs/contracts/pilot/language-and-tone.md

@@ -56,7 +56,7 @@ Examples + wrong-vs-correct: [`language-and-tone-examples`](../../docs/guideline
 
 ## `.md` files — ALWAYS English
 
-Every text inside `.md` under `.augment/`, `.agent-src/`, `.agent-src.uncompressed/`, `agents/`: headings, paragraphs, bullets, option labels, prompts, placeholders, ASCII labels, table headers / content. Agent translates at runtime.
+Every text inside `.md` under `.augment/`, `.agent-src/`, `.agent-src.uncondensed/`, `agents/`: headings, paragraphs, bullets, option labels, prompts, placeholders, ASCII labels, table headers / content. Agent translates at runtime.
 
 **Labeled-anchor exception** — quoting German inside English prose is forbidden. Either translate, OR use a labeled `DE: … · EN: …` anchor block (only allowed location for German prose).
 

package/docs/contracts/plain-language-surface.md

@@ -0,0 +1,117 @@
+---
+stability: beta
+keep-beta-until: 2026-08-26
+roadmap_ref: road-to-frictionless-employee-workspace.md
+---
+
+# Plain-Language Surface Contract
+
+> **Status** · v0 / beta · 2026-05-26. Defines the user-facing
+> relabel matrix for non-developer audiences. Phase C of
+> `road-to-frictionless-employee-workspace.md`. The matrix is the
+> single source of truth; lint enforcement lives in
+> `scripts/lint_role_experiences.py --plain-language`.
+
+## What this is
+
+A relabel matrix between the package's **technical vocabulary**
+(council, trust, packs, orchestration, contracts, advisory, …) and
+the **plain-language labels** non-developer audiences (galabau owner,
+sales rep, support agent, team leader, content creator, consultant)
+expect to read.
+
+The matrix does **not** rename the technical vocabulary in
+developer-facing artefacts (`SKILL.md`, rule files, ADRs,
+`docs/contracts/`, `docs/decisions/`, `agents/contexts/`, scripts).
+The technical vocabulary earned its place there. The matrix only
+governs three user-facing surfaces:
+
+1. **Workspace UI strings** (`src/ui/pages/WorkspacePage.tsx` and its
+   child components — buttons, headings, tooltips, banner text).
+2. **`docs/getting-started-by-role.md`** — the non-developer entry
+   surface. Engineering-lead and developer sections keep the
+   technical vocabulary.
+3. **Role experience `index.md` bodies** (`agents/roles/<slug>/index.md`)
+   — the one-paragraph identity + first-task descriptions. The
+   `skills.yml` files are link-targets, not user prose, and keep
+   the skill IDs verbatim.
+
+## Why
+
+The 9.3/10 feedback round (2026-05-25) caught six jargon terms a
+sales rep / galabau owner / support agent does not parse on first
+read: `council`, `trust level`, `pack`, `orchestration`, `contract`,
+`advisory`. Replacing them in the non-developer surface lifts the
+"Daily workspace UX" score from 5.5/10 toward the ≥ 7.5/10 acceptance
+gate; the technical vocabulary survives in the developer-facing
+artefacts where reviewers, contributors, and agents continue to read
+them.
+
+## Matrix
+
+| Technical term | Plain-language label | Applies to surface |
+|---|---|---|
+| `council` | `second-opinion check` | Workspace UI · role indices · getting-started |
+| `trust level` | `reliability score` | Workspace UI · role indices · getting-started |
+| `pack` | `ready-made setup` | Workspace UI · role indices · getting-started |
+| `orchestration` | `multi-step workflow` | Workspace UI · role indices · getting-started |
+| `contract` | `guarantee` | Workspace UI · role indices · getting-started |
+| `advisory` | `recommendation` | Workspace UI · role indices · getting-started |
+
+## Carve-outs — surfaces that keep the technical vocabulary
+
+The matrix MUST NOT apply to:
+
+- `.agent-src.uncondensed/`, `.agent-src/`, `.augment/`, `.claude/`,
+  `.cursor/`, `.windsurf/`, `.clinerules/` — all skill / rule /
+  command / persona artefacts.
+- `docs/contracts/`, `docs/decisions/`, `docs/architecture.md`,
+  `docs/getting-started-laravel.md`, every other developer-facing
+  doc.
+- `agents/contexts/`, `agents/roadmaps/`, `scripts/`, source code,
+  test fixtures, golden output files.
+- The developer / engineering-lead sections inside
+  `docs/getting-started-by-role.md` — those readers are the audience
+  for the technical vocabulary.
+- `agents/roles/<slug>/skills.yml` — the skill IDs are URL slugs,
+  not user prose. The `why:` strings stay neutral so the launcher's
+  shortlist rail is informative; both audiences read them.
+
+## Lint enforcement
+
+`scripts/lint_role_experiences.py --plain-language` scans every
+`agents/roles/<slug>/index.md` body for the six jargon terms.
+Frontmatter and fenced code blocks are excluded. Any hit fails the
+lint with a pointer back to this contract. Phase C Step 5 wires the
+flag into `task ci`.
+
+## Tone scaffold
+
+When a relabel feels awkward, default to the role's own language:
+the galabau owner does not call something a "second-opinion check"
+verbatim either, but they understand it instantly. Compare:
+
+- **Technical:** "the council ruled this trust level is `community`."
+- **Plain:** "the second-opinion check rated this reliability score
+  as community-built (less reviewed than core)."
+
+The plain version is longer; that is the cost of readability for
+non-developer audiences. Workspace UI strings stay short by carrying
+the relabel without the explanation (the explanation lives in the
+plain-explain right-rail toggle).
+
+## Re-audit cadence
+
+Re-audit on each of:
+- A new jargon term entering the user-facing surface (caught either
+  by lint or by a feedback round).
+- A role experience adding a section the matrix does not cover.
+- A reviewer round (recruit-session) that names a relabel the matrix
+  missed.
+
+## See also
+
+- [`docs/getting-started-by-role.md`](../getting-started-by-role.md) — the entry surface.
+- [`docs/contracts/role-experience.md`](role-experience.md) — role artefact shape.
+- [`scripts/lint_role_experiences.py`](../../scripts/lint_role_experiences.py) — lint enforcement.
+- `road-to-frictionless-employee-workspace.md` § Phase C — the roadmap step.

package/docs/contracts/profile-system.md

@@ -18,7 +18,7 @@ keep-beta-until: 2026-08-14
 
 A **profile** declares the user's audience identity. Six seed profiles
 ship; users can declare their own under
-`.agent-src.uncompressed/profiles/<id>.yml`.
+`.agent-src.uncondensed/profiles/<id>.yml`.
 
 | `profile.id` | Audience | README entry-paragraph | Default `preset.id` |
 |---|---|---|---|
@@ -101,7 +101,7 @@ command (Phase 1 item 3).
 ## User-defined profiles
 
 A consumer project MAY ship a custom profile under
-`.agent-src.uncompressed/profiles/<id>.yml`. Constraints:
+`.agent-src.uncondensed/profiles/<id>.yml`. Constraints:
 
 - `id` MUST be unique across seed + user-defined profiles.
 - Shape MUST match the seed contract above (audience / defaults / surface).
@@ -121,7 +121,7 @@ Only changes to the **seed set** require an ADR.
 - A profile YAML references a non-existent `preset_id`.
 - The seed-profile count diverges from this contract's table.
 - `defaults.personas` references a persona id that does not exist
-  under `.agent-src.uncompressed/personas/`.
+  under `.agent-src.uncondensed/personas/`.
 
 ## Non-goals
 

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

@@ -0,0 +1,103 @@
+# Release-PR Gating Contract
+
+> **Status:** active · **Owner:** maintainer (`scripts/release.py`) · **Opened:** 2026-05-26
+>
+> Source: `road-to-optimized-ci-and-release-gates.md` Phase A Step 1. Measured
+> baseline (`gh run list --branch main --limit 50`): `Public Install Smoke`
+> avg **413 s** (3-OS × 2-Node matrix), `Tests` avg **218 s** (Linux + macOS
+> + Windows). Both trigger on `package.json`. Release PRs (`release/X.Y.Z`)
+> only touch `package.json`, `CHANGELOG.md`, `marketplace.json`,
+> `packages/*/pack.yaml`, `packages/*/README.md` — verified against PR #238
+> (3.3.0). They cannot regress install or runtime behaviour by construction.
+
+## Release-PR shape
+
+A pull request qualifies as a **release PR** when **both** of the following
+hold:
+
+1. **Head branch matches** `^release/\d+\.\d+\.\d+$` — same regex as
+   `scripts/release.py` § `_RELEASE_BRANCH_RE`.
+2. **Diff file set is a subset of the version-bump allowlist:**
+   - `package.json`
+   - `CHANGELOG.md`
+   - `.claude-plugin/marketplace.json`
+   - `packages/*/pack.yaml`
+   - `packages/*/README.md`
+
+Both predicates are enforced by `scripts/check_release_pr_shape.py`
+(stdlib-only, ≤ 150 LOC). The script exits 0 when both hold; non-zero with
+a per-file diff naming any out-of-allowlist entry otherwise.
+
+## Cut surface — heavy jobs that skip on release PRs
+
+Skipped via `if: !startsWith(github.head_ref, 'release/')` guards on the
+heavy install/test jobs. These are the jobs that release PRs cannot regress
+by construction (no install scripts, no runtime code, no test source in the
+release-PR allowlist):
+
+| Workflow | Job | Why it cuts |
+|---|---|---|
+| `tests.yml` | `install-tests` | release-PR diff has no `install.sh` / `scripts/install.py` / `tests/test_install.sh` |
+| `tests.yml` | `install-aux-tests` | same — orchestrator, key contracts, one-liner smoke all untouched |
+| `tests.yml` | `python-tests` | release-PR diff has no `scripts/**` or `tests/**` (other than CHANGELOG via path filter — see below) |
+| `tests.yml` | `node-tests` | release-PR diff has no `src/**`, `tests/{cli,server,ui}/**`, `packages/core/installer/**` |
+| `tests.yml` | `windows-lockfile-export` | release-PR diff has no `scripts/install_global*.py`, `scripts/cmd_export.py`, lockfile test surface |
+| `smoke-public-install.yml` | `smoke` | release-PR diff has no `scripts/install*`, `setup.sh`, `templates/**`, `package.json` runtime behaviour |
+
+`push:` to `main` and the weekly cron on `smoke-public-install.yml` stay
+**unconditional** — those catch drift the PR matrix can't see.
+
+## Kept surface — release PRs still prove these
+
+The release-PR required-check floor stays equivalent (smaller, faster) to
+the feature-PR floor by adding:
+
+| Workflow | Job | Proves |
+|---|---|---|
+| `consistency.yml` | (existing) | `task consistency` — source-of-truth integrity |
+| `smoke.yml` | `smoke-contracts` | Contract self-checks (kernel, router, hashes) |
+| `release-guard.yml` | `assert-version-matches-tag` | already gates `npm publish`; remains tag-trigger |
+| `migration-dry-run.yml` | (existing) | Migration plan dry-runs |
+| `release-validation.yml` (Phase B) | `release-shape` | shape detector — fails closed if diff exits the allowlist |
+| `release-validation.yml` (Phase B) | `changelog-entry` | CHANGELOG carries an entry matching the head-branch version |
+| `release-validation.yml` (Phase B) | `version-consistency` | `package.json` / `marketplace.json` / pack manifests agree on the version |
+
+## Rollback trigger — fail-closed
+
+The optimisation is **opt-in by shape, not by branch name**. A release-PR
+whose diff contains a stray file outside the allowlist (e.g. a last-minute
+CHANGELOG fixup that also touches `scripts/release.py`) trips the shape
+detector:
+
+1. `check_release_pr_shape.py` exits non-zero with a per-file diff.
+2. The CI dashboard surfaces "release-shape" red.
+3. The maintainer either narrows the diff (move the script edit to a
+   separate PR) or accepts the heavy matrix re-running on the next push.
+
+In other words: the cut applies only when shape is **provably** safe. Branch
+name alone never bypasses the heavy matrix.
+
+## What this contract is not
+
+- **Not a test-deletion contract.** Every existing assertion still runs
+  somewhere in the cadence (PR for feature PRs, push-to-main + weekly cron
+  for smoke).
+- **Not a coverage-reduction contract.** The release-validation jobs are
+  additive, not substitutive.
+- **Not a release-velocity contract.** Release cadence is driven by
+  Conventional Commits in `scripts/release.py`, not by CI cost.
+- **Not a Hard-Floor lift.** No security check is removed.
+  `release-guard.yml`'s `assert-version-matches-tag` job is independent of
+  `Tests` / `Public Install Smoke` and stays mandatory on every tag.
+
+## See also
+
+- `docs/contracts/branch-protection-policy.md` — per-PR-shape required-check
+  matrix (Phase D).
+- `docs/contracts/ci-cost-budget.md` — measured baselines + quarterly review
+  cadence (Phase C).
+- `.github/workflows/release-validation.yml` — the tight release-shape
+  validation workflow (Phase B).
+- `scripts/check_release_pr_shape.py` — the shape detector (Phase A).
+- `scripts/release.py` § `_RELEASE_BRANCH_RE` — source of truth for the
+  release-branch naming convention.

package/docs/contracts/role-experience.md

@@ -1,6 +1,6 @@
 ---
 stability: beta
-keep-beta-until: 2026-08-24
+keep-beta-until: 2026-08-23
 ---
 
 # Role-experience contract
@@ -19,7 +19,7 @@ A folder at `agents/roles/<role>/` that contains:
 
 A role experience is **not**:
 
-- A new skill — the shortlist references existing skills under `packages/<pack>/.agent-src.uncompressed/skills/`. The role experience curates; it does not implement.
+- A new skill — the shortlist references existing skills under `packages/<pack>/.agent-src.uncondensed/skills/`. The role experience curates; it does not implement.
 - A documentation page — `docs/getting-started-by-role.md` is the prose surface and links **to** the role experience, never duplicates it.
 - A persona file under `personas/` — personas are review voices the AI Council uses; role experiences are end-user entry points.
 
@@ -92,7 +92,7 @@ skills:
     why: "Section-by-section drafting flow for longer offers."
 ```
 
-Every `id` must resolve to an existing skill under `packages/<pack>/.agent-src.uncompressed/skills/<id>/SKILL.md` — verified by the lint pass in Phase 3 Step 6.
+Every `id` must resolve to an existing skill under `packages/<pack>/.agent-src.uncondensed/skills/<id>/SKILL.md` — verified by the lint pass in Phase 3 Step 6.
 
 ## Versioning + status
 

package/docs/contracts/rule-classification.md

@@ -7,7 +7,7 @@ keep-beta-until: 2026-08-12
 # Rule Classification — Phase 1.2 of road-to-kernel-and-router
 
 > **Status:** active · **Stability:** beta · **Owner:** road-to-kernel-and-router P1.2
-> · **Source of truth:** `.agent-src.uncompressed/rules/*.md`
+> · **Source of truth:** `.agent-src.uncondensed/rules/*.md`
 
 Migration plan for all 56 source rules. **No edits land here** — this
 document is the disposition map P2 / P4 will execute against.
@@ -16,8 +16,8 @@ document is the disposition map P2 / P4 will execute against.
 
 | code | meaning |
 |---|---|
-| `keep-in-kernel` | Iron Law floor / behaviour / safety / tone / ask-policy. Loaded every session. Compressed to ≤ 2.5k chars (P2.2; raised from 1.5k per Council R2). |
-| `compress-and-keep` | Behavioural rule that stays in `rules/` as auto-tier. Compressed in place per the P2.2 playbook. Loaded by router on trigger match. |
+| `keep-in-kernel` | Iron Law floor / behaviour / safety / tone / ask-policy. Loaded every session. Condensed to ≤ 2.5k chars (P2.2; raised from 1.5k per Council R2). |
+| `condense-and-keep` | Behavioural rule that stays in `rules/` as auto-tier. Condensed in place per the P2.2 playbook. Loaded by router on trigger match. |
 | `move-to-skill:<id>` | Procedural how-to content. Body migrates into the named skill (P4.1); rule shrinks to Iron Law one-liner + `routes_to:`. |
 | `move-to-guideline:<id>` | Reference / examples / mechanics tables. Body migrates into `docs/guidelines/<id>.md` (P4.2); rule keeps Iron Law + pointer. |
 
@@ -51,25 +51,25 @@ yet.
 | scope-control | 4368 | `keep-in-kernel` | Iron Law (safety-floor): no unsolicited refactors / git-ops gate |
 | verify-before-complete | 2344 | `keep-in-kernel` | Iron Law: no completion claims without fresh verification |
 
-**Always-bucket total: 32 403 chars.** Pilot compression rate `r`
+**Always-bucket total: 32 403 chars.** Pilot condensation rate `r`
 locked at **median 0.712** (P1.3, Council R2 amendment); projection
 sum = 23 071 chars, under the 25k hard cap with 1 929 chars headroom.
 Per-rule cap raised to ≤ 2.5k (Council R2). Iron-Law-override ADRs
 may lift individual rules above 2.5k where Iron-Law density forbids
-further compression — currently 2 expected (`direct-answers` +342,
+further condensation — currently 2 expected (`direct-answers` +342,
 `language-and-tone` +1286). See `kernel-membership.md` § 5.
 
 > **Council swap ADR (P2.1 input).** Sonnet 4.5 flagged
 > `agent-authority` as a routing index (no Iron-Law fence; should
 > demote to auto-tier-3) and `autonomous-execution` (currently
-> `compress-and-keep` below) as a mode-independent Band-4 authority
+> `condense-and-keep` below) as a mode-independent Band-4 authority
 > that should promote to kernel. Swap accepted ⇒ +1213 chars over
 > the 25k cap. Decision deferred to P2.1 ADR; current dispositions
 > below remain locked until the ADR resolves.
 
-### § 3.2 — auto-rules: compress-and-keep (22)
+### § 3.2 — auto-rules: condense-and-keep (22)
 
-Behavioural rules with too much Iron-Law content for migration; compress
+Behavioural rules with too much Iron-Law content for migration; condense
 in place per P2.2 playbook. Loaded by router on trigger match.
 
 | id | chars | tier | rationale |
@@ -101,7 +101,7 @@ in place per P2.2 playbook. Loaded by router on trigger match.
 
 Procedural how-to. Rule body migrates into named skill (P4.1); rule
 shrinks to Iron-Law one-liner + `routes_to:`. Skill IDs reference
-existing skills under `.agent-src.uncompressed/skills/`.
+existing skills under `.agent-src.uncondensed/skills/`.
 
 | id | chars | target skill | rationale |
 |---|---:|---|---|
@@ -130,7 +130,7 @@ Reference / examples / mechanics. Body migrates into
 
 | id | chars | target guideline | rationale |
 |---|---:|---|---|
-| artifact-engagement-recording | 3462 | `artifact-engagement-flow` | Mechanics doc lives in `.agent-src.uncompressed/contexts/contracts/` (P4.1) |
+| artifact-engagement-recording | 3462 | `artifact-engagement-flow` | Mechanics doc lives in `.agent-src.uncondensed/contexts/contracts/` (P4.1) |
 | augment-portability | 2956 | `augment-portability-patterns` | Project-agnostic-patterns reference |
 | command-suggestion-policy | 3954 | `command-suggestion-flow` | Flow doc already exists |
 | php-coding | 3433 | `php-coding-patterns` | Reference table; per-stack |
@@ -142,9 +142,9 @@ Reference / examples / mechanics. Body migrates into
 
 | check | command | acceptance |
 |---|---|---|
-| Coverage | `python3 scripts/measure_rule_budget.py` | 56 rules total → 9 kernel + 22 compress + 18 skill + 7 guideline |
+| Coverage | `python3 scripts/measure_rule_budget.py` | 56 rules total → 9 kernel + 22 condense + 18 skill + 7 guideline |
 | Determinism | `python3 scripts/measure_rule_budget.py --json` × 2 | byte-identical output |
-| No-edits | `git diff --stat .agent-src.uncompressed/rules/` | clean (this is plan-only) |
+| No-edits | `git diff --stat .agent-src.uncondensed/rules/` | clean (this is plan-only) |
 
 ## § 5 — Open questions
 
@@ -166,6 +166,6 @@ Reference / examples / mechanics. Body migrates into
    spec; flagged by Council R2 to prevent inconsistent implementer
    calls.
 5. **`think-before-action` / `context-hygiene` / `augment-source-of-
-   truth` boundary.** Currently `compress-and-keep` (auto). Council
+   truth` boundary.** Currently `condense-and-keep` (auto). Council
    did not flag for promotion; revisit only if the swap ADR (#1)
    forces a re-projection.

package/docs/contracts/rule-interactions.md

@@ -6,7 +6,7 @@ keep-beta-until: 2026-08-12
 # Rule-Interaction Matrix
 
 > **Audience:** rule authors and reviewers — anyone editing
-> `.agent-src.uncompressed/rules/*.md` or proposing a new always-rule.
+> `.agent-src.uncondensed/rules/*.md` or proposing a new always-rule.
 > **Authoritative source:** [`rule-interactions.yml`](rule-interactions.yml).
 > **Linter:** `scripts/lint_rule_interactions.py` (run via `task lint-rule-interactions`).
 
@@ -73,8 +73,8 @@ The YAML uses six relation kinds. Definitions:
   conflict: …                                                  # what triggers both
   resolution: …                                                # what the agent does
   evidence:
-    - .agent-src.uncompressed/rules/non-destructive-by-default.md#the-iron-law
-    - .agent-src.uncompressed/rules/autonomous-execution.md#hard-floor--see-non-destructive-by-default
+    - .agent-src.uncondensed/rules/non-destructive-by-default.md#the-iron-law
+    - .agent-src.uncondensed/rules/autonomous-execution.md#hard-floor--see-non-destructive-by-default
 ```
 
 `rules: [a, b]` is ordered: `a` is senior (wins on conflict), `b` is
@@ -117,7 +117,7 @@ source of truth and weaken it. Canonical contracts:
 | Decision-trace shape (what the engine emits per phase) | [`decision-trace-v1.md`](decision-trace-v1.md) |
 | Memory contract (entries, scopes, retention) | [`agent-memory-contract.md`](agent-memory-contract.md) |
 | Memory visibility in the trace (`affected` keys) | [`memory-visibility-v1.md`](memory-visibility-v1.md) |
-| AI-Council consultation flow | [`../skills/ai-council/SKILL.md`](../../.agent-src.uncompressed/skills/ai-council/SKILL.md) |
+| AI-Council consultation flow | [`../skills/ai-council/SKILL.md`](../../.agent-src.uncondensed/skills/ai-council/SKILL.md) |
 
 Where an `always`-rule **does** interact with one of these surfaces
 (e.g. `non-destructive-by-default` gating a memory-driven action), the