@sabaiway/agent-workflow-memory 1.1.1 → 1.2.0

package/CHANGELOG.md

@@ -4,6 +4,39 @@ All notable changes to the memory substrate. Versions are this **package's** npm
 they are distinct from the **deployment-lineage** stamp written into a project's
 `docs/ai/.memory-version` (which tracks the shared `agent-workflow` lineage, head `1.3.0`).
 
+## 1.2.0 — Seeds the per-project orchestration config
+
+The substrate now seeds a new **per-project, user-editable recipe config** —
+`references/templates/orchestration.json` — deployed into `docs/ai/orchestration.json` by the bootstrap
+template loop. It declares the orchestration **recipe** each named activity/slot uses (the composition
+root's read-only `procedures` advisor reads it); the recipe **canon** and the slot **vocabulary** still
+live in the engine / composition root, never here (the substrate keeps knowing nobody — the seed's
+self-documentation uses generic "composition root" phrasing, naming no sibling). The shipped default is
+conservative: **`solo` everywhere**, with an onboarding `_README` explaining how to raise a slot.
+
+### Added
+- **`references/templates/orchestration.json`** — strict JSON (no comments), byte-identical to the kit's
+  fallback copy (kit↔memory template parity guard). Seeded on bootstrap; on upgrade it is **ensured
+  stamp-independently** (create-if-missing / **preserve-if-edited** — a user's edits are never clobbered),
+  so even an equal-head (`1.3.0`) deployment gains it **without a lineage-head bump or a migration file**.
+- An **ownership-table** row distinguishing the seeded, editable recipe **CONFIG** (memory) from the
+  recipe **CANON** + slot vocabulary (engine / composition root).
+
+The deployment-lineage head stays **`1.3.0`** (no `docs/ai` structural change; no migration file). The
+npm package version is a separate axis.
+
+## 1.1.2 — Entry-point template headroom for the orchestration pointer
+
+A **docs/prose** release (no new executable, the `1.1.1`/`1.9.1` precedent). The bundled entry-point
+template (`references/templates/AGENTS.md`) ships a second empty marker pair — `workflow:orchestration`,
+right under the methodology pair — which the family **composition root** fills live from the methodology
+engine on deploy. To keep the deployed `AGENTS.md` inside its ≤100-line budget with **both** pointers
+filled, the template trimmed non-essential slack (the Hard-Constraints intro blockquote + one
+illustrative row a deploying agent adapts anyway). No behaviour change; the composition root remains the
+only writer of the slots.
+
+The deployment-lineage head stays **`1.3.0`** (no `docs/ai` structural change; no migration file).
+
 ## 1.1.1 — Installer hardening (Issue-004 parity)
 
 A patch release that applies the same two installer fixes shipped to the engine in 1.1.0, keeping the

package/README.md

@@ -33,10 +33,11 @@ Open your agent inside a project and run the skill:
 - **`/agent-workflow-memory`** — bootstrap a new or empty project. Asks the three setup
   questions (visibility, conversational language, agent attribution), then writes `docs/ai/`,
   `AGENTS.md`, the enforcement scripts + pre-commit hook, and stamps
-  `docs/ai/.memory-version`. The methodology slot in `AGENTS.md` is left **empty**.
+  `docs/ai/.memory-version`. The two pointer slots in `AGENTS.md` (the **methodology** and
+  **orchestration** pointers, filled later by the family composition root) are left **empty**.
 - **`/agent-workflow-memory upgrade`** — migrate an existing deployment to the skill's current
   version. Reads `docs/ai/.memory-version`, runs only newer migrations in semver order, and
-  **preserves** anything already in the methodology slot (extract-and-reinsert — it never
+  **preserves** anything already in either pointer slot (extract-and-reinsert — it never
   regenerates `AGENTS.md` wholesale).
 
 ## Stamps & lineage

package/SKILL.md

@@ -3,7 +3,7 @@ name: agent-workflow-memory
 description: Deploy or upgrade a portable AI-agent memory substrate in any project — an entry-point `AGENTS.md` (+ `CLAUDE.md` alias) and a structured `docs/ai/` context store with cap/archive/index enforcement. Use when the user wants to bootstrap `docs/ai/`, set up the Memory Map and session protocols, install the docs-rotation pre-commit hook, or run `/agent-workflow-memory` / `/agent-workflow-memory upgrade`. Triggers on "set up the memory system", "deploy the AI memory here", "bootstrap docs/ai", "upgrade the memory substrate". This is the substrate only — the workflow methodology (plan→execute→review, queue, Cleanup) is owned elsewhere and injected into AGENTS.md by the family composition root.
 disable-model-invocation: true
 metadata:
-  version: '1.1.1'
+  version: '1.2.0'
 ---
 
 # agent-workflow-memory
@@ -20,11 +20,12 @@ manually).
 
 This skill is the **memory layer** of the `agent-workflow` family. It **knows nobody else** in
 the family. In particular it does **not** own the **workflow methodology** (plan → execute →
-review vocabulary, lifecycle, `docs/plans/queue.md`, mandatory Cleanup, plan-then-execute) —
-that is injected into a delimited slot in `AGENTS.md` by the family **composition root**, never by
-this skill. (This substrate does not name or depend on any specific sibling — it only honours the
-slot contract.) This skill only ever ships the slot **empty** and **preserves** whatever is already
-in it on upgrade.
+review vocabulary, lifecycle, `docs/plans/queue.md`, mandatory Cleanup, plan-then-execute) **or the
+orchestration recipes** — those are injected into **two delimited pointer slots** in `AGENTS.md`
+(`workflow:methodology` + `workflow:orchestration`) by the family **composition root**, never by this
+skill. (This substrate does not name or depend on any specific sibling — it only honours the slot
+contract.) This skill only ever ships **both** slots **empty** and **preserves** whatever is already
+in them on upgrade.
 
 The substrate **artifacts** (this skill, the templates, the deployed `docs/ai/` files) stay in
 their **source language** — for cross-agent and cross-team portability. That is separate from the
@@ -35,8 +36,8 @@ contents).
 
 ## Ownership table
 
-What this substrate owns vs what it only points at. The methodology pointer is the empty slot
-the composition root fills — never author methodology text here.
+What this substrate owns vs what it only points at. The methodology + orchestration pointers are the
+**two empty slots** the composition root fills — never author that text here.
 
 | Concern | Owner | In the deployed `AGENTS.md` |
 |---|---|---|
@@ -45,6 +46,8 @@ the composition root fills — never author methodology text here.
 | Visibility / conversational-language / agent-attribution contracts | **memory** | the three `AGENTS.md` blocks |
 | Deployment-lineage stamp | **memory** | `docs/ai/.memory-version` |
 | Plan→Phase→Step vocabulary, lifecycle, `queue.md`, mandatory Cleanup | **methodology** (not this skill) | the empty `workflow:methodology` slot — filled by the composition root |
+| Orchestration recipes (Solo / Reviewed / Council / Delegated) | **methodology engine** (not this skill) | the empty `workflow:orchestration` slot — filled by the composition root |
+| Per-project recipe **CONFIG** (which recipe each activity/slot uses) | **memory** seeds an *editable default* | `docs/ai/orchestration.json` (hand-edited; the recipe **canon** + the slot **vocabulary** live in the engine / composition root, never here) |
 
 ---
 
@@ -90,10 +93,12 @@ bootstrapping over a live system, but the user makes the final call.
 5. **Entry-point doc.** If `AGENTS.md` / `CLAUDE.md` already exist (recon), do **not**
    overwrite — show the user and ask whether to merge or replace. Otherwise create `AGENTS.md`
    from `${CLAUDE_SKILL_DIR}/references/templates/AGENTS.md` and symlink `CLAUDE.md -> AGENTS.md`
-   (`ln -s AGENTS.md CLAUDE.md`). **Leave the `workflow:methodology` slot exactly as shipped —
-   empty.** Filling it is the composition root's job.
+   (`ln -s AGENTS.md CLAUDE.md`). **Leave BOTH pointer slots (`workflow:methodology` +
+   `workflow:orchestration`) exactly as shipped — empty.** Filling them is the composition root's job.
 6. **Deploy `docs/ai/`.** Create the files + `pages/` from
-   `${CLAUDE_SKILL_DIR}/references/templates/`. Keep each file's frontmatter.
+   `${CLAUDE_SKILL_DIR}/references/templates/` (every non-`AGENTS.md` template, including the seeded,
+   **user-editable** `docs/ai/orchestration.json` config — strict JSON, the per-project recipe defaults
+   the composition root's `procedures` advisor reads). Keep each file's frontmatter.
 7. **Fill templates** per the table below.
 8. **Install enforcement (Node projects).** Copy `${CLAUDE_SKILL_DIR}/references/scripts/*.mjs`
    (+ `*.test.mjs`) into the project's `scripts/`. **No Node runtime** → skip this + the hook;
@@ -108,7 +113,7 @@ bootstrapping over a live system, but the user makes the final call.
 10. **Stamp the deployment lineage.** Write the **deployment-lineage head** into
     `docs/ai/.memory-version` (one semver line). The lineage head is **`1.3.0`** (the
     `LINEAGE_HEAD` constant in `scripts/stamp-takeover.mjs`) — the shared `agent-workflow`
-    lineage, **not** this package's npm version (`1.1.0`). Use the atomic writer in
+    lineage, **not** this package's npm version (`1.2.0`). Use the atomic writer in
     `scripts/stamp-takeover.mjs` (write-temp + rename).
 11. **Report & ask.** Show `tree docs/ai/`, 2–3 lines on filled-vs-TODO, then **ask before
     committing** — never auto-commit. **Exception — delegated mode (below): skip this gate.**
@@ -116,8 +121,9 @@ bootstrapping over a live system, but the user makes the final call.
 > **Delegated mode (invoked by the composition root) — applies to BOTH bootstrap and upgrade.**
 > When the family composition root drives this substrate as part of a family bootstrap **or
 > upgrade**, do the write steps (bootstrap 1–10 / upgrade 1–7: write `docs/ai/` + `AGENTS.md` +
-> `.memory-version`) but **do NOT** run **any** commit gate and **do NOT** ask to commit — the
-> composition root owns the **single** commit gate, raised after it injects the methodology slot.
+> `.memory-version`, **including seeding / stamp-independently ensuring `docs/ai/orchestration.json`**)
+> but **do NOT** run **any** commit gate and **do NOT** ask to commit — the
+> composition root owns the **single** commit gate, raised after it injects the two pointer slots.
 > The three setup answers and the target dir are passed in by the composition root; you perform no
 > commit and no slot injection. **Standalone** invocation keeps its own commit gate (bootstrap step
 > 11 / upgrade step 7).
@@ -154,8 +160,14 @@ Fill strategy:
    block and removes it only after the user's explicit consent — never by default); if it is
    **untracked AND not ignored** → **AMBIGUOUS** → **ASK** the user before writing. This visibility check
    runs on **every** in-range upgrade, even at head — it is not gated by the stamp delta, but it is gated
-   **behind** the never-downgrade STOP above. **Then**, if the stamp **equals** the head → report "up to
-   date" (plus any footprint move just made) and stop.
+   **behind** the never-downgrade STOP above. **Also stamp-independent (same gate, before the equal-head
+   short-circuit): ensure `docs/ai/orchestration.json`** — **create it from
+   `${CLAUDE_SKILL_DIR}/references/templates/orchestration.json` if missing**, **preserve it byte-for-byte
+   if it already exists** (a user may have edited it; never clobber it). This is why an equal-head (`1.3.0`)
+   deployment still gains the config seed **without a lineage-head bump or a migration file** (the
+   stamp-independent-reconcile precedent — like the pointer slots + the hidden-mode footprint). **Then**,
+   if the stamp **equals** the head → report "up to date" (plus any footprint move / config seed just
+   made) and stop.
 3. Show the relevant `${CLAUDE_SKILL_DIR}/CHANGELOG.md` context (entries newer than the stamp).
 4. Apply `${CLAUDE_SKILL_DIR}/migrations/<version>-<slug>.md` in **semver order**, only those
    newer than the stamp. Migrations are **idempotent**.
@@ -164,12 +176,12 @@ Fill strategy:
    pre-1.1.0 deployment with no *Communication language* block, ask + insert
    (`migrations/1.1.0-communication-language.md`); pre-1.2.0 with no *Attribution* block, ask +
    insert defaulting to `off` (`migrations/1.2.0-agent-attribution.md`).
-6. **Preserve the methodology slot.** If `AGENTS.md` has the `workflow:methodology` markers,
-   **never regenerate the file wholesale** — extract any bytes between the markers and reinsert
-   them unchanged. If the markers are absent (a legacy `AGENTS.md`), gracefully **no-op** on the
-   slot (adding the slot to already-deployed files is a separate methodology migration). On any
-   malformed marker state (single, reversed, nested, or duplicate pair), **no-op with an error**
-   — never edit.
+6. **Preserve BOTH pointer slots.** If `AGENTS.md` has the `workflow:methodology` and/or
+   `workflow:orchestration` markers, **never regenerate the file wholesale** — extract any bytes
+   between each pair and reinsert them unchanged. If a pair is absent (a legacy `AGENTS.md`),
+   gracefully **no-op** on that slot (adding a slot to already-deployed files is the composition
+   root's reconcile, not this substrate's job). On any malformed marker state (single, reversed,
+   nested, or duplicate pair), **no-op with an error** — never edit.
 7. **Re-stamp** `docs/ai/.memory-version` to the lineage head (atomic write). Report changes;
    **ask before committing** — **except in delegated mode** (see the *Delegated mode* note above),
    where the composition root owns the single gate and this step raises none.
@@ -181,9 +193,10 @@ Fill strategy:
 - **Source vs target directory.** Templates/scripts are read from the skill's own dir; the
   **working directory is the target project** — never write substrate files back into the skill.
 - **Stamp = lineage head, not package version.** `.memory-version` carries `1.3.0` (the shared
-  `agent-workflow` lineage), not the npm `1.1.0`. They are independent axes.
-- **The methodology slot ships empty and stays the user's.** Never author methodology text into
-  it; on upgrade, preserve its content byte-for-byte. The composition root is its only writer.
+  `agent-workflow` lineage), not the npm `1.2.0`. They are independent axes.
+- **Both pointer slots ship empty and stay the user's.** Never author methodology or orchestration
+  text into them; on upgrade, preserve their content byte-for-byte. The composition root is their only
+  writer.
 - **The `Co-Authored-By` trailer is added by the harness, not by prose.** When attribution is
   `off` + Claude Code, also set `"includeCoAuthoredBy": false` in `.claude/settings.json`.
 - **Hidden mode is project-local, never machine-global.** Memory's hide writes to the repo's
@@ -209,8 +222,9 @@ The three setup choices each have a full contract in
 ## References
 
 - [`references/contracts.md`](references/contracts.md) — the three setup contracts in full.
-- [`references/templates/`](references/templates/) — stack-agnostic `AGENTS.md` (with the empty
-  methodology slot), `agent_rules.md`, and all `docs/ai/` files to deploy.
+- [`references/templates/`](references/templates/) — stack-agnostic `AGENTS.md` (with the two empty
+  pointer slots — methodology + orchestration), `agent_rules.md`, the seeded user-editable
+  `orchestration.json` config, and all `docs/ai/` files to deploy.
 - [`references/scripts/`](references/scripts/) — the Node enforcement scripts (caps + staleness +
   index-freshness gate, 3-tier archive, hook installer) and their unit tests.
 - [`scripts/stamp-takeover.mjs`](scripts/stamp-takeover.mjs) — the upgrade-time lineage state

package/bin/install.test.mjs

@@ -33,6 +33,7 @@ describe('memory installer — payload + symlink-traversal hardening', () => {
       'scripts/stamp-takeover.mjs',
       'references/contracts.md',
       'references/templates/AGENTS.md',
+      'references/templates/orchestration.json',
       'migrations/legacy-stamp-takeover.md',
     ]) {
       assert.ok(existsSync(join(target, f)), `missing installed entry: ${f}`);

package/capability.json

@@ -3,7 +3,7 @@
   "schema": 1,
   "name": "agent-workflow-memory",
   "kind": "memory-substrate",
-  "version": "1.1.1",
+  "version": "1.2.0",
   "provides": ["context"],
   "roles": {},
   "detect": {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@sabaiway/agent-workflow-memory",
-  "version": "1.1.1",
+  "version": "1.2.0",
   "description": "Portable, cross-agent memory substrate for AI coding agents — an AGENTS.md entry point + docs/ai context with cap/archive/index enforcement, deployable standalone or as part of the agent-workflow family. The memory layer of the agent-workflow family.",
   "keywords": [
     "ai-agents",

package/references/scripts/check-docs-size.mjs

@@ -139,7 +139,10 @@ export const parseStaleAfter = (value) => {
   return Number(m[1]);
 };
 
-const walkMarkdownFiles = async (dir) => {
+// Discover the docs to validate: ONLY `*.md` files (recursively). Non-`.md` files — e.g. a hand-edited
+// `docs/ai/orchestration.json` config — are inherently skipped, so they are never subject to the
+// frontmatter / maxLines caps. Exported so that skip is pinned by a regression test.
+export const walkMarkdownFiles = async (dir) => {
   const entries = await readdir(dir, { withFileTypes: true });
   const files = [];
   for (const entry of entries) {

package/references/scripts/check-docs-size.test.mjs

@@ -10,6 +10,7 @@ import {
   inspectFile,
   buildIndex,
   checkIndexFreshness,
+  walkMarkdownFiles,
 } from './check-docs-size.mjs';
 
 describe('parseFrontmatter', () => {
@@ -126,6 +127,29 @@ describe('inspectFile', () => {
   });
 });
 
+// The cap-validator discovers ONLY `*.md` files, so a non-`.md` doc — e.g. a hand-edited
+// `docs/ai/orchestration.json` config — is inherently skipped: never validated for frontmatter / caps.
+// This belt-and-suspenders regression pins that skip so adding a config `.json` under docs/ai can never
+// start failing the docs gate.
+describe('walkMarkdownFiles — only *.md is discovered (a config .json is skipped)', () => {
+  let dir;
+  beforeEach(async () => {
+    dir = await mkdtemp(join(tmpdir(), 'walk-md-test-'));
+  });
+  afterEach(async () => {
+    await rm(dir, { recursive: true, force: true });
+  });
+
+  it('returns the .md file and NOT a sibling orchestration.json', async () => {
+    await writeFile(join(dir, 'doc.md'), '---\ntype: reference\nmaxLines: 100\n---\n\nbody.\n');
+    await writeFile(join(dir, 'orchestration.json'), '{ "plan-authoring": { "review": "reviewed" } }\n');
+    const found = await walkMarkdownFiles(dir);
+    expect(found.some((f) => f.endsWith('doc.md'))).toBe(true);
+    expect(found.some((f) => f.endsWith('.json'))).toBe(false);
+    expect(found.some((f) => f.endsWith('orchestration.json'))).toBe(false);
+  });
+});
+
 // Synthetic row matching the shape produced by `inspectFile` + `formatRow`.
 const makeRow = (path, overrides = {}) => ({
   path,

package/references/templates/AGENTS.md

@@ -55,19 +55,18 @@ Start-of-session, during-work, and task-completion procedures live in [`docs/ai/
 
 <!-- workflow:methodology:start -->
 <!-- workflow:methodology:end -->
+<!-- workflow:orchestration:start -->
+<!-- workflow:orchestration:end -->
 
 ---
 
 ## 🚫 Hard Constraints
 
-> Adapt to this stack — remove rows that don't apply. Style rules should be linter-enforced, not prose.
-
 | Rule | Enforcement |
 |------|-------------|
 | No `export default` (named exports only) | Linter |
 | No `any` / no unsafe casts | Linter / type-checker |
 | Functional style (no mutation in app code) | Linter |
-| No single-letter variables | Code review |
 | Interactive elements semantic (button/link, not div+onClick) | Linter / a11y |
 | No hardcoded colors — design tokens only | Code review |
 | No business logic in components → hooks/services | Architecture review |

package/references/templates/orchestration.json

@@ -0,0 +1,5 @@
+{
+  "_README": "Per-project orchestration config: the recipe used at each step (slot) of each named activity. Hand-edit this file — it is never written for you. Each activity is configured independently (e.g. plan-authoring, plan-execution), and so is each slot within it. A slot's value is a recipe: a 'review' slot accepts solo | reviewed | council (you self-review / one backend reviews / both review and you synthesize); an 'execute' slot accepts solo | delegated (you implement / a backend runs a bounded sub-task). The default below is 'solo' everywhere — no execution backend required. Raise a slot to reviewed or council for a second opinion, or to delegated to hand off execution; those need an execution backend set up first. Remove a slot's line to fall back to the computed default (reviewed when a review backend is ready, otherwise solo). Run the read-only procedures advisor to see an activity's steps plus the recipe resolved for your environment, and pass a per-run override to change one slot just once. Strict JSON — no comments.",
+  "plan-authoring": { "review": "solo" },
+  "plan-execution": { "execute": "solo", "review": "solo" }
+}

package/scripts/standalone-bootstrap.test.mjs

@@ -18,6 +18,7 @@ import {
   cpSync,
   readdirSync,
   readFileSync,
+  writeFileSync,
   existsSync,
   symlinkSync,
 } from 'node:fs';
@@ -35,6 +36,17 @@ const ENFORCEMENT = join(SKILL_ROOT, 'references', 'scripts');
 // from the composition root) so this substrate test stays self-contained and dependency-free.
 const SLOT_START = '<!-- workflow:methodology:start -->';
 const SLOT_END = '<!-- workflow:methodology:end -->';
+const ORCH_START = '<!-- workflow:orchestration:start -->';
+const ORCH_END = '<!-- workflow:orchestration:end -->';
+// The deployed AGENTS.md line budget the composition root fills BOTH pointers inside (D-CAP). A
+// representative single-line fragment per slot models what the composition root injects.
+const AGENTS_MD_CAP = 100;
+const lineCount = (text) => text.split('\n').length - (text.endsWith('\n') ? 1 : 0);
+const extractPair = (text, start, end) => {
+  const a = text.indexOf(start);
+  const b = text.indexOf(end);
+  return a !== -1 && b !== -1 && b > a ? text.slice(a + start.length, b) : null;
+};
 
 const tempDirs = [];
 const makeProject = () => {
@@ -98,14 +110,71 @@ describe('standalone substrate bootstrap (end-to-end, real temp project)', () =>
     assert.ok(!existsSync(join(docsAi, '.workflow-version')), 'no composition-root stamp in a standalone bootstrap');
   });
 
-  it('ships the methodology slot present-but-empty', () => {
+  it('ships BOTH pointer slots present-but-empty (methodology + orchestration)', () => {
     const project = makeProject();
     bootstrap(project);
 
     const entry = readFileSync(join(project, 'AGENTS.md'), 'utf8');
-    const start = entry.indexOf(SLOT_START);
-    const end = entry.indexOf(SLOT_END);
-    assert.ok(start !== -1 && end !== -1 && end > start, 'an ordered slot marker pair is present');
-    assert.equal(entry.slice(start + SLOT_START.length, end).trim(), '', 'the slot is empty as shipped');
+    const meth = extractPair(entry, SLOT_START, SLOT_END);
+    const orch = extractPair(entry, ORCH_START, ORCH_END);
+    assert.notEqual(meth, null, 'an ordered methodology marker pair is present');
+    assert.equal(meth.trim(), '', 'the methodology slot is empty as shipped');
+    assert.notEqual(orch, null, 'an ordered orchestration marker pair is present');
+    assert.equal(orch.trim(), '', 'the orchestration slot is empty as shipped');
+  });
+
+  it('seeds docs/ai/orchestration.json from the template (the bootstrap loop deploys it)', () => {
+    const project = makeProject();
+    const docsAi = bootstrap(project);
+    const seeded = join(docsAi, 'orchestration.json');
+    assert.ok(existsSync(seeded), 'the orchestration.json config is seeded into docs/ai');
+    assert.equal(
+      readFileSync(seeded, 'utf8'),
+      readFileSync(join(TEMPLATES, 'orchestration.json'), 'utf8'),
+      'the seeded config is byte-identical to the template',
+    );
+    // strict JSON valid + the conservative all-solo default the maintainer chose.
+    const config = JSON.parse(readFileSync(seeded, 'utf8'));
+    assert.equal(typeof config._README, 'string', 'an onboarding _README is present');
+    assert.equal(config['plan-authoring'].review, 'solo', 'default review recipe is solo');
+  });
+
+  // The stamp-independent upgrade "ensure" (SKILL.md upgrade step 2): create-if-missing /
+  // preserve-if-edited. Modeled here the way the documented prose performs it — so an equal-head
+  // re-run never clobbers a user's edited config, and a deleted one is re-seeded.
+  it('the upgrade ensure preserves an edited config and re-creates a deleted one', () => {
+    const project = makeProject();
+    const docsAi = bootstrap(project);
+    const dest = join(docsAi, 'orchestration.json');
+    const ensureConfig = () => {
+      if (!existsSync(dest)) cpSync(join(TEMPLATES, 'orchestration.json'), dest);
+    };
+
+    // A user edits the deployed config.
+    writeFileSync(dest, '{ "plan-authoring": { "review": "council" } }\n');
+    ensureConfig(); // an equal-head upgrade re-runs the ensure
+    assert.match(readFileSync(dest, 'utf8'), /council/, 'an edited config is preserved (never clobbered)');
+
+    // A missing config is re-seeded.
+    rmSync(dest);
+    ensureConfig();
+    assert.ok(existsSync(dest), 'a missing config is re-created from the template');
+    assert.equal(readFileSync(dest, 'utf8'), readFileSync(join(TEMPLATES, 'orchestration.json'), 'utf8'));
+  });
+
+  it('stays ≤ the cap when the composition root fills BOTH pointer slots (D-CAP headroom)', () => {
+    const project = makeProject();
+    bootstrap(project);
+    const entry = readFileSync(join(project, 'AGENTS.md'), 'utf8');
+    assert.ok(lineCount(entry) <= AGENTS_MD_CAP, `shipped (empty) AGENTS.md is ${lineCount(entry)} lines (cap ${AGENTS_MD_CAP})`);
+    // Fill each one-line pointer the way the composition root does (replace the empty body), then re-count.
+    const fill = (text, start, end, body) => {
+      const a = text.indexOf(start);
+      const b = text.indexOf(end);
+      return `${text.slice(0, a + start.length)}\n${body}\n${text.slice(b)}`;
+    };
+    let filled = fill(entry, SLOT_START, SLOT_END, '> methodology pointer (one line)');
+    filled = fill(filled, ORCH_START, ORCH_END, '> orchestration recipes pointer (one line)');
+    assert.ok(lineCount(filled) <= AGENTS_MD_CAP, `dual-filled AGENTS.md is ${lineCount(filled)} lines (cap ${AGENTS_MD_CAP})`);
   });
 });