mindforge-cc 11.9.0 → 11.9.2

package/CHANGELOG.md

@@ -1,5 +1,219 @@
 # Changelog
 
+## [11.9.2] — 2026-08-16 — Correctness: audit-chain integrity, dashboard crash policy, secret scanning
+
+Patch release. No new features. Correctness work closing defects found by a
+multi-agent audit of v11.9.1, plus the regression suites that keep them closed.
+Contains a breaking change to the dashboard HTTP surface — see BREAKING below.
+
+### BREAKING
+
+Shipped under a PATCH bump. The break is confined to the dashboard's own HTTP
+surface, which binds to 127.0.0.1 only — but if you script against it, read this.
+
+- **Dashboard error responses changed shape.** `detail` is removed from 5 endpoints and
+  raw errno strings from 10 more; `correlation_id` is added to 15; a malformed request
+  body now returns `application/json` instead of express's `text/html` error page.
+  Anything parsing `detail` must correlate on the logged `correlation_id` instead. This
+  was deliberate — those fields leaked absolute filesystem paths, and therefore the
+  operator's username and home directory, into an unauthenticated response body
+  (`requireAuth` exempts GET).
+- **The dashboard now EXITS on an unhandled rejection or uncaught exception** where
+  11.9.1 logged and continued. If you supervise the process, expect restarts where you
+  previously saw a logged error. Rationale in the Fixed section below: log-and-continue
+  held client sockets open until the client timed out, and had made `shutdown()` swallow
+  a throwing token unlink and keep serving the authenticated mutation API after SIGTERM.
+- **`node bin/validate-config.js` and `mindforge security-scan` can now fail.** They
+  previously reported `MINDFORGE.md valid — 0 settings configured` and exited 0 on every
+  input. If you run either in CI, a genuinely invalid registry will now red-line where it
+  used to pass. Note this reaches **fresh installs and `--force` reinstalls only** — the
+  installer does not overwrite an existing `.mindforge/MINDFORGE-SCHEMA.json`, so a plain
+  upgrade keeps the old permissive schema.
+
+### Fixed
+
+- **`security-scan` could not fail.** `bin/validate-config.js` and
+  `bin/models/model-router.js` each parsed `MINDFORGE.md` with a plain `KEY=value` regex, but
+  the registry declares its 43 parameters as bracketed `[KEY] = value`. Every schema property
+  resolved to `undefined` and short-circuited, so the command reported
+  `MINDFORGE.md valid — 0 settings configured` and exited 0 on any input. The schema also had
+  no `required` key at all. Both parsers now share `bin/utils/mindforge-params.js`, which also
+  accepts the legacy plain form (`examples/starter-project/MINDFORGE.md` ships 28 such lines),
+  and the schema declares real `required`/`recommended` sets.
+  **Behaviour change for consumers:** three CI gates go from unfailable to failable —
+  `.github/workflows/mindforge-ci.yml:38`, `.gitlab-ci-mindforge.yml:12`, and
+  `.github/workflows/control-plane.yml:100`. If one red-lines on a valid value, the schema
+  bound is wrong; do not "fix" it by editing `MINDFORGE.md`. Model routing is unchanged —
+  30 persona x tier combinations resolve identically.
+- **Trace retrieval returned nothing usable.** Queries were wrapped as a single FTS phrase, so
+  any query containing one absent term scored zero; and `traces_search` was keyed on `trace_id`
+  rather than the primary key, so each span's DELETE evicted the previous span and only the last
+  span per trace stayed searchable — 2,270 of 5,117 content-bearing traces, 44.4%, unsearchable.
+  Queries are now tokenised, OR-joined and ranked by tf-idf (`matchinfo('pcnx')`); the index is
+  re-keyed and rebuilt losslessly from the base table. `bin/eval/eval-harness.js` and its golden
+  set had zero callers and are now reachable as `npm run eval:retrieval`, with the baseline
+  committed: mean recall@10 0.6417, nDCG 0.5698 over 519 documents.
+- **The cost ledger reported two totals for one concept.** `sum(cost_usd)` was $13.73 while
+  `sum(total_cost_usd)` was $0.00, and `tests/dashboard.test.js` wrote the reader's field name,
+  so the mismatch tested green. `bin/models/usage-record.js` is now the single definition of the
+  ledger path, record shape, per-entry cost and day bucket. The configured `ledger_path` pointed
+  at `token-ledger.jsonl`, a file that has never existed; that ghost filename had spread to 17
+  places across 13 files and is now absent. The dashboard cost tile no longer renders `$0.00`
+  on a 500 — it had no `res.ok` check, and because errors return well-formed JSON the catch
+  never fired, making an outage indistinguishable from zero spend.
+  A maintainer tool, `scripts/purge-synthetic-usage.js`, removes fixture rows: dry-run by
+  default, backs up first, idempotent, and aborts leaving the ledger untouched if the rewrite
+  fails. It is run from a repository checkout — `scripts/` is not in the published tarball, so
+  installed consumers do not have it.
+- **Audit hash chain could fork under concurrent writers.** `bin/autonomous/audit-writer.js`
+  read the chain head and appended with no mutual exclusion, and cached the head in-process
+  indefinitely — so once a second process appended, the first kept chaining from a superseded
+  hash. Added `bin/utils/file-lock.js` (a fail-closed advisory lock promoted from
+  `bin/learning/instinct-cli.js`, deliberately NOT from `.agent/bin/lib/state.cjs`, which
+  writes anyway when the lock cannot be taken) held across read-head-through-fsync, and made
+  the cached head carry the file size that witnesses it is still the tail. 8 concurrent
+  appenders went from 199 broken links + 4 forks per 200 entries to 0. A lock alone was
+  measured insufficient — it still left 2 breaks and 1 fork, because the stale cache is a
+  second, independent defect.
+- **Knowledge-graph edge updates were lost under concurrency.** `deprecateEdge`,
+  `reinforceEdge` and `applyDecay` in `bin/memory/knowledge-graph.js` each did
+  `readAllEdges()` -> mutate -> append with no lock, and `addEdge` appended unserialised
+  against them. Measured at HEAD over 4 runs of 8 processes x 20 `reinforceEdge` calls:
+  93-129 of 160 increments lost, final `traversal_count` 31-67 instead of 160. All four
+  write paths now hold the `graph-edges.jsonl` lock across read-through-append; the same
+  probe then loses 0 of 160 in every run, with no lock-acquisition failures.
+- `bin/governance/policy-engine.js`: `logAudit`'s un-awaited audit write now has a
+  `.catch()` — a lock-contention failure is reported at the decision site instead of
+  escaping as an unhandled rejection.
+- `bin/hooks/instinct-capture-hook.js` appended to the instinct store without the lock that
+  `instinct-cli`'s prune/import rewrite holds, so a hook append landing in that window was
+  clobbered by the rename. It now takes the same lock.
+- `tests/v7-sovereign-security.test.js`: `new PolicyEngine()` no longer defaults
+  `planningDir` to `process.cwd()`, which appended test verdicts to the operator's real
+  `.planning/RISK-AUDIT.jsonl`.
+- Packaging: `package.json` files[] now excludes `**/*.lock` so a lockfile orphaned by a
+  hard kill cannot leak into the npm tarball (verified: without the negation, a
+  `.mindforge/memory/graph-edges.jsonl.lock` does ship).
+
+- **Audit-chain forgery via un-awaited rollback.** `HindsightInjector.inject` called
+  the async `TemporalHub.rollbackTo` without `await`, so its rejection escaped the
+  surrounding `try/catch` while execution continued: a failed rollback still fsync'd a
+  hash-chained `hindsight_injected` entry and flipped `auto-state.json` to
+  `awaiting_regeneration` for something that never happened. The log gained a
+  cryptographically valid record of a non-event, and `verify-audit` reported the chain
+  valid — valid and wrong.
+- **CLI `defaultArgs` were replaced by user arguments, not prepended.** `mindforge
+  health <anything>` lost `--check` and fell through to `installer-core`'s real
+  `install()` with `force: true`. Now prepended.
+- **`_verifyMetadata` compared UTF-16 code units, not bytes**, so a 64-unit / 65-byte
+  `integrity` still threw `RangeError` — and the caller degraded that throw into
+  "proceeding without integrity check" and restored the snapshot anyway. Read and
+  verify are now separate stages; only a genuinely absent `SNAPSHOT-META.json` reaches
+  the tolerant path.
+- **Dashboard `RevOpsAPI` was required but never mounted**, so `/api/revops/overview`
+  404'd while the AgRevOps panels and docs described it as live. Even mounted it threw:
+  `getAuditEntries()` returns `{entries,total,limit,offset}` and three engines call
+  `.filter()` on it.
+- **Dashboard leaked error internals to clients.** `err.message` reached response
+  bodies from 14 sites (`api-router.js` ×10, `temporal-api.js` ×3, `revops-api.js` ×1);
+  for fs-sourced errors that string carries absolute paths, disclosing the operator's
+  username and home directory. All sites now log server-side and return a generic
+  message plus a `correlation_id`, behind a 4-arg terminal handler that stops express
+  rendering `err.stack` when `NODE_ENV !== 'production'`.
+- **Test runner discovery was a flat `readdirSync`**, so any suite in a subdirectory was
+  invisible. Now a recursive walk that prunes `tmp-*` / `node_modules` / dot
+  directories — directories only, never files.
+- Three orphan files removed: one truncated `.planning/AUDIT.jsonl` to zero, one
+  overwrote `.planning/STATE.md`, one called a function absent from `bin/`. Recursion
+  made the `STATE.md` clobberer reachable by a single in-place rename; the other two
+  were unreachable by the runner at any depth. **The claim in `5177225`'s message that
+  all three were newly armed by recursion is correct for one of the three and
+  over-attributed for the other two.**
+- Three relocated demos kept one-level-up requires after moving a directory deeper, so
+  all three exited 1; corrected to `../../bin/`.
+
+### Changed
+
+- **`mindforge audit-skill`, `register-skill`, `install-skill` and `record-learning` no
+  longer carry `defaultArgs`.** Prepending turned them from inert into live state
+  writers: `audit-skill <name> <ver> <tier>` appended a hash-chained
+  `{event:'skill_installed', validation_passed:true}` entry for a skill that does not
+  exist, and `register-skill` wrote a malformed row above the table header of
+  `.mindforge/org/skills/MANIFEST.md`, which ships in the tarball. Bare invocations now
+  print usage and exit 0 without reaching those writers, as at v11.9.1.
+- **`subagent` is a first-class command.** Prepending `spawn` shadowed `spawn-agent`'s
+  `subagent` mode, whose documented route was `mindforge spawn subagent <name>`.
+- **The dashboard now exits on an unhandled rejection instead of logging and
+  continuing.** An escaped rejection is the only reliable signal that an async call was
+  left un-awaited, and express 4.22.1 does not route async handler rejections to error
+  middleware — log-and-continue held the client socket open until the client gave up
+  (2.5s, 4s and 8s clients all timed out) versus a ~15ms reset on exit. Symmetric with
+  `uncaughtException`, whose log-and-continue form had made `shutdown()` swallow a
+  throwing token unlink and keep serving the token-authenticated mutation API **after
+  SIGTERM**, with the bearer token still on disk and valid in memory.
+- **Dashboard error responses changed shape.** `detail` is removed from 5 endpoints,
+  raw errno strings from 10 more, `correlation_id` is added to 15, and a malformed
+  request body now returns `application/json` rather than express's `text/html` error
+  page. Anything parsing `detail` must correlate on the logged id instead.
+
+### Added
+
+- **Secret scanning enforced at three layers**: `.gitleaks.toml`, a `.husky/pre-commit`
+  gate that fails loudly when gitleaks is absent rather than skipping, and
+  `.github/workflows/secret-scan.yml` scanning full history. `scripts/ci/verify-secret-scan.sh`
+  self-tests the scanner — it distinguishes "scanned clean" from "scanned and found"
+  from "did not scan", because gitleaks exits 1 for both a finding and a failed config
+  load, and writes no report in the latter case.
+
+### Tests
+
+- Suite totals for this release: **105 files, 103 pass, 2 environment-dependent skips**
+  (`browser`, `sre-integration`). Eight new suites across the release:
+  `temporal-integrity`, `dashboard-error-leak`, `dashboard-crash-guards`,
+  `dashboard-wiring`, `cli-router`, `mindforge-params`, `file-lock`, `retrieval-fts`.
+- **Four suites could not report failure and now can.** `v8-persistence`,
+  `v8-skill-evolution` and `v8-orbital-governance` ended `finally { process.exit(0) }`, and
+  `v7-pillar-integration` had zero assertions with a premium-model gate that named two models
+  absent from the registry for several releases. `npm test` is the only quality step before
+  `npm publish`, and the runner gates on child exit codes, so a blind suite blinded the publish
+  gate. Verified by injected failure rather than inspection.
+  `dashboard-wiring` derives the expected router set
+  from `server.js`'s own requires, so adding a router without mounting it fails.
+  `cli-router` runs against a mirror-root sandbox under `os.tmpdir()` — required, not
+  tidiness: the case that proves audit forgery is prevented would otherwise forge an
+  entry into the real chain on every run. `revops-roi.test.js` had 0 assertions and
+  could not fail; it now has 6.
+
+### Not fixed — deferred to v12
+
+- **No hook is registered in any consumer install.** The installer copies 9 hook
+  scripts into `<runtime>/hooks/`, but nothing writes `.claude/settings.json` and it is
+  absent from `package.json` `files[]`. Verified by installing the tarball into a
+  scratch project. Every gate this release hardens is inert until that lands.
+- **`requireAuth` exempts GET and OPTIONS**, so every read route — including
+  `/api/audit`, which serves the hash-chained audit log — is credential-free to any
+  local process. Mutations are protected. This is a threat-model decision, not a patch.
+- **`audit-skill audit <name> <ver> <tier>`** — the explicit form — still reaches a
+  writer that performs no existence check and hardcodes `validation_passed: true`. Only
+  the bare invocation is closed.
+- **Snapshot integrity is not an authenticity control.** `HMAC_KEY` is a literal in
+  shipped source, the HMAC covers only the metadata object so file **contents** are
+  unsigned (editing a file inside a signed snapshot leaves the signature valid), and
+  deleting `SNAPSHOT-META.json` bypasses verification entirely.
+- `cwd: ROOT` in the CLI, which resolves consumer state inside `node_modules`.
+- `security-scan` cannot fail: its parser expects `KEY=value` while `MINDFORGE.md` uses
+  `[KEY] = value`, so it always reports 0 settings and exits 0.
+- Version drift in six publishable manifests (`Formula/mindforge.rb`, `Dockerfile`,
+  `mcp-server/server.json`, `mcp-server/src/index.ts`, the plugin manifest and the
+  marketplace entry) is untouched here — none is gated, and the Formula pins a tarball
+  sha256 that cannot exist before publish.
+
+## [11.9.1] — 2026-07-29 — Packaging Fix: Restore Missing Workflow Commands
+
+### Fixed
+- A blanket `.claude/` rule in `.gitignore` (pre-existing, unrelated to the v11.9.0 work) was silently excluding 34 command files under `.claude/commands/mindforge/` from git — mostly the `wf-*` dynamic-workflow slash commands, plus `skill-tdd.md`, `skills-index.md`, and `systematic-debug.md`. Since `package.json`'s `files` array includes `.claude/commands/`, this meant **v11.9.0's published npm tarball was missing every `wf-*` slash command** even though the mirrored `.agent/mindforge/` copies shipped correctly. Removed the redundant gitignore rule (the repo's existing `*.local` pattern already covers the actual intended exclusion, `.claude/settings.local.json`) and committed all 34 files.
+- Added a regression guard to `tests/packaging-allowlist.test.js` asserting `.claude/commands/mindforge/` and `.agent/mindforge/` ship the same file count — the prior `>=150` floor check passed throughout this gap's entire lifetime since both counts already exceeded 150 independently.
+
 ## [11.9.0] — 2026-07-27 — Bedrock Provider + Full Dry-Run Audit
 
 ### Added
@@ -220,1825 +434,6 @@ CI governance-gate gap, and tightens two security surfaces. No new features.
   `federated-sync` un-loadable in a clean install. Swapped to the built-in
   `crypto.randomUUID()` (zero-native-deps); both modules now load.
 
-## [11.5.0] - 2026-06-11 — Governance hardening + autonomous-engine repair (Waves 4–7)
-
-This release bundles four waves of work: orchestration primitives, an **inert** manifest
-engine + the new instinct CLI + GAN harness personas, governance/security hardening, and a
-repair that makes the autonomous engine (`/mindforge:auto`) actually functional. Several
-items ship deliberately **inert** (scaffolding present, not wired into any live path) and
-are flagged as such below — they introduce **no behavior change yet**.
-
-### Added
-
-- **Instinct CLI** (`/mindforge:instinct`, `bin/learning/instinct-cli.js`) — deterministic,
-  no-LLM management of the JSONL instinct store: `list`, `export`, `import`,
-  `promote-candidates` (list + flag), and `prune`.
-- **Typed Inter-Agent Message Protocol** (`bin/autonomous/handoff-schema.js`) — five message
-  kinds (`task_handoff`, `query`, `response`, `completed`, `conflict`) with priority levels
-  (`low`/`normal`/`high`/`critical`) and message validation. Internal orchestration primitive.
-- **Manifest-driven install resolver** (`bin/installer/install-manifests.js`,
-  `.mindforge/manifests/install-*.json`) — profile-to-module expansion and dependency
-  detection. Ships **INERT**: the adapter wiring it into `bin/installer-core.js`'s live
-  `install()` path is deferred to a future PR. No behavior change.
-- **GAN-style harness personas** (`.mindforge/personas/gan-evaluator.md`, `gan-generator.md`,
-  `gan-planner.md`) — fully scoped and documented. Ships **INERT**: not wired to any live
-  command or automated workflow. No behavior change.
-- **Governance test coverage** — `tests/trust-verifier.test.js` (7 tests) and
-  `tests/rbac-manager.test.js` (8 tests) lock the fail-closed contracts for future
-  safety-critical governance changes.
-
-### Fixed
-
-- **SSRF import-URL guard** (`bin/learning/lib/ssrf-guard.js`) — closed an IPv6 link-local
-  bypass (`fe81`–`fe8f` reachable via string-prefix check) and a symlink path-traversal
-  bypass, using numeric bitmask validation and canonical-path checking before system-dir
-  validation.
-- **Federated EIS sync** (`bin/memory/eis-client.js`) — `getAuthHeader` no longer throws on
-  every call; it now registers a node identity via `ZTAI.registerAgent` and signs with
-  `ZTAI.signData`, restoring sync to non-localhost EIS endpoints.
-- **RBAC tier elevation** (`bin/governance/rbac-manager.js`) — `getRolesByTier` now fails
-  safely for unregistered agents (no thrown exception) and resolves agent tier via the
-  correct ZTAI API.
-- **Autonomous engine — wave crash** (`bin/autonomous/auto-runner.js`) — replaced ZTAI
-  singleton misuse (`getIdentity()`, non-existent) with `_getRunnerIdentity()` using the
-  real `registerAgent` API; `/mindforge:auto` no longer crashes on every wave.
-- **Autonomous policy gate — fail-open** (`bin/autonomous/auto-runner.js`) — the async
-  policy verdict is now `await`ed (`this.policyEngine.evaluate(intent)`); the per-wave gate
-  previously always allowed.
-- **Autonomous engine — fail-closed identity** (`bin/autonomous/auto-runner.js`) — if runner
-  identity cannot be established, the policy gate now denies and audits (`auto_mode_denied`)
-  instead of proceeding ungoverned.
-
-### Changed
-
-- **Instinct store schema** (`.mindforge/engine/instincts/instinct-schema.md`) — added
-  `project_id` (stable scope key) and `source` (`auto-capture`/`manual`/`imported`/
-  `observer`), plus origin-based confidence scoring: auto-capture starts at `0.3`, manual
-  at `0.7`.
-- **Cost-routing shadow mode** (`bin/revops/router-steering-v2.js`) — arbitrage steering now
-  respects `cost_routing.shadow_mode` (default `true`); in observe-only mode `steer()`
-  returns `{ shadow, authoritative: false }` with SHADOW vs LIVE logging.
-
-### Security
-
-- **Tier-3 approvals now fail closed** (`bin/governance/approve.js`) — approvals require GPG
-  verification and **throw before writing any record** when no GPG key is configured, unless
-  `MINDFORGE_ALLOW_UNVERIFIED_APPROVAL=1` is set; unverified approvals are marked
-  `verified: false`.
-- **Persona supply-chain scan** (`scripts/ci/validate-assets.js`) — persona asset validation
-  now detects dangerous invisible unicode (zero-width, bidi overrides, Unicode tags) across
-  `.mindforge/personas` to prevent ASCII-smuggling injection.
-- **Destructive-command detector** (`bin/security/trust-boundaries.js`) — `normalizeShell`
-  now strips bare `#` tokens after quote-stripping, blocking quoted-hash evasion (e.g.
-  `rm "#" -rf /`).
-- **Instinct import port allowlist** (`bin/learning/lib/ssrf-guard.js`) — `validateImportUrl`
-  enforces an `ALLOWED_IMPORT_PORTS` allowlist (`{'', '443'}`), blocking attempts to reach
-  internal services (Redis `:6379`, Mongo `:27017`, etc.) on otherwise-allowed public hosts.
-
-## [11.4.0] - 2026-06-06 — Claude Code plugin distribution
-
-MindForge is now installable as a native **Claude Code plugin** from a marketplace, in
-addition to the `npx mindforge-cc` installer. This release ships the plugin tooling so
-both the GitHub plugin channel and the npm tarball stay coherent.
-
-### Added
-
-- **Plugin marketplace** (`.claude-plugin/marketplace.json` at the repo root) — users run
-  `/plugin marketplace add sairam0424/MindForge` then `/plugin install mindforge@mindforge`.
-  Lists 11 plugins: the comprehensive `mindforge` plugin + 10 à-la-carte subagent packs.
-- **Comprehensive `mindforge` plugin** (`plugins/mindforge/`) bundling the full surface:
-  174 commands, 154 subagents, 73 skills + a synthesized `mindforge-protocol` skill (the
-  CLAUDE.md operating directive, since a plugin-root CLAUDE.md isn't loaded as context),
-  and governance hooks (translated to Claude `PreToolUse`/`PostToolUse` events with
-  `${CLAUDE_PLUGIN_ROOT}`-relative paths).
-- **Bundled MindForge MCP server** — a self-contained esbuild single-file build
-  (`mcp/dist/index.js`, all deps inlined, no runtime `node_modules`) exposing 7 tools over
-  stdio: `mindforge_health`, `mindforge_status`, `mindforge_memory_query`,
-  `mindforge_memory_stats`, `mindforge_memory_find_related`, `mindforge_audit_log`
-  (read-only), and append-only `mindforge_memory_remember`. Scoped to the user's project
-  via `${CLAUDE_PROJECT_DIR}`; degrades gracefully when MindForge isn't set up.
-- **Generators** (single source of truth, drift-guarded): `build-plugin-marketplace.js`,
-  `build-subagent-plugins.js`, `build-mindforge-plugin.js`, `vendor-sdk-into-mcp.js`,
-  `fix-command-frontmatter.js`; new `mcp-server/` package with esbuild build.
-- **`tests/plugin-packaging.test.js`** (in `npm test`) — guards the generated plugin tree,
-  frontmatter validity, and MCP-bundle self-containment.
-
-### Fixed
-
-- Quoted YAML-unsafe `name`/`description` frontmatter across commands/agents/skills that
-  the plugin validator flagged (leading `-`/`@`, embedded `: `, trailing `:`) — these
-  silently loaded empty metadata. The npx installer inherits the fix.
-- Per-category subagent `plugin.json` `agents[]` now derived from on-disk files, fixing
-  stale bare names for the 16 collision-renamed (`-cc`) agents.
-
-## [11.3.1] - 2026-06-05 — Packaging hotfix (commands, skills & framework now ship)
-
-Critical fix for v11.3.0, where a too-narrow npm `files` allowlist silently dropped
-most of the product from the published tarball. Users who ran `npx mindforge-cc` got
-only hooks, personas, subagents, and three `.mindforge/` folders — **no slash commands,
-no skills, and an incomplete framework** — with no error (the installer skips any source
-absent from the tarball). v11.3.1 restores the full payload and adds a regression test
-so this class of bug cannot ship silently again.
-
-### Fixed
-
-- **npm `files` allowlist** (`package.json`) widened to ship what the installer reads:
-  `.claude/commands/` (174 slash commands), `.agent/mindforge/` + `.agent/forge/`,
-  `.agent/skills/` (73 skills), the entry `.claude/CLAUDE.md`, and the full `.mindforge/`
-  framework (`governance`, `integrations`, `intelligence`, `memory`, `metrics`, `models`,
-  `org`, `plugins`, `team` — previously only `engine`/`personas`/`skills` reached users).
-  Runtime state is explicitly **excluded** (`celestial.db`, `metrics/token-usage.jsonl`,
-  `memory/pattern-library.jsonl`) — `files` overrides `.npmignore`, so these are negated
-  in the allowlist itself.
-- **`.planning/` scaffolding now ships and installs** — sourced from the clean, generic
-  `examples/starter-project/.planning/` (never the framework repo's own live `.planning/`,
-  which holds dev state: `AUDIT.jsonl`, `slack-threads.json`, `jira-sync.json`). Added the
-  four missing starter templates (`ROADMAP.md`, `REQUIREMENTS.md`, `ARCHITECTURE.md`,
-  `RELEASE-CHECKLIST.md`) so the autonomous engine (`/mindforge:auto`, `:next`) has the
-  files it expects.
-- **`docs/References` / `docs/Templates` case-sensitivity bug** — the installer read
-  lowercase `docs/references`/`templates`, which "worked" on case-insensitive macOS but
-  silently missed on case-sensitive Linux/npm (manifest showed `REFERENCES 0 / TEMPLATES 0`).
-  Installer now reads the exact on-disk case and both dirs are allowlisted.
-
-### Added
-
-- **`tests/packaging-allowlist.test.js`** — packs the real tarball (`npm pack --dry-run
-  --json`) and asserts commands, skills, subagents, entry files, the full `.mindforge/`
-  framework, and `.planning/` templates all ship, while runtime DBs/telemetry do not.
-  Proven to fail under the broken v11.3.0 allowlist. Skips loudly if npm is unavailable.
-
-## [11.3.0] - 2026-06-04 — "Legion" (154-subagent expansion)
-
-Imports the full VoltAgent `awesome-claude-code-subagents` collection (154 Claude-Code-native
-subagents across 10 categories) into MindForge, fully rebranded. Additive and
-backward-compatible — no existing persona, skill, or command changed behavior.
-
-### Added
-
-- **`subagents/` tree** — 154 specialized subagent definitions in 10 categories
-  (`01-core-development` … `10-research-analysis`), kept verbatim from upstream except
-  the rebrand and collision/functional fixes below. Agent bodies are unmodified prose.
-- **`.mindforge/imported-agents.jsonl`** — generated index (name → path → category → model →
-  description) and **`scripts/build-subagent-index.js`** to regenerate it deterministically.
-- **`bin/spawn-agent.js subagent <name>`** — third loader mode that resolves an imported
-  subagent via the index. Hardened with a strict name allowlist (`[A-Za-z0-9-_]`), exact
-  index-match resolution (never builds a path from input), and a `subagents/` containment
-  check — defense-in-depth path-traversal guards consistent with `BLOCK_ON_SECURITY`.
-- **Installer delivery** — `bin/installer-core.js` now installs the 154 subagents into the
-  runtime's native agents directory (`.claude/agents/`, via the new `agentsSubdir` runtime
-  config + `installSubagents()`), flattening the category tree (READMEs excluded) so Claude
-  Code auto-discovers them. Runs for both global and local scope; mirrored to `.claude/agents/`
-  for non-claude local runtimes. Shown in `--dry-run` and the payload manifest; `--uninstall`
-  removes only the imported set, never the user's own agents. The `.mindforge/imported-agents.jsonl`
-  index ships with `.mindforge/`, so the `subagent` loader mode also works post-install.
-- **`tests/subagent-import.test.js`** — asserts 154 indexed, every path exists, all 16
-  collisions renamed, no bare-name clash with personas, and the loader's traversal guards.
-
-### Changed (rebrand — VoltAgent → MindForge)
-
-- Plugin/marketplace metadata (`subagents/.claude-plugin/marketplace.json` + 10 category
-  `plugin.json`) renamed `voltagent-*` → `mindforge-*`, author → MindForge Team, URLs →
-  `github.com/sairam0424/MindForge`. Per-plugin versions preserved for upstream re-sync.
-  No live marketplace is published — metadata only.
-- 16 name collisions with existing hand-authored MindForge personas resolved with a `-cc`
-  suffix (`api-designer-cc`, `debugger-cc`, …); the original personas remain authoritative
-  for the bare names.
-- **agent-installer** fetch URLs repointed to `sairam0424/MindForge` under `subagents/`.
-- **design-bridge** VoltAgent design-md dependency replaced with a `<DESIGN_MD_SOURCE>`
-  placeholder + a needs-configuration notice (MindForge ships no design-md repo yet).
-- Dead VoltAgent Discord/sponsor links removed (no MindForge equivalent).
-- `package.json` `files` allowlist now ships `subagents/` + the index; README capability
-  line notes the 154 subagents.
-
-## [11.2.1] - 2026-05-31 — "Hardening" (security & integrity audit remediation)
-
-Post-v11.2.0 audit remediation. Closes every exploitable security defect and
-false-assurance stub surfaced by the end-to-end audit. No new features, no
-breaking changes — fixes and honest-labeling only.
-
-### Security (fixed)
-
-- **trust-gate-hook**: scans the whole command + every line (a benign first line could previously cloak a destructive later line)
-- **orbital-guardian**: `verify()` re-checks the Ed25519 attestation signature (added `did`/`signed_message` columns + migration); rejects forged APPROVED rows
-- **policy-engine**: `reasoning_proof` alone no longer bypasses the blast-radius limit (`isProofValid` inits false; cryptographic `pq_proof` path unchanged)
-- **shadow-mirror**: git calls use `execFileSync` (argv) + fail-closed `sanitizeRemediationId()` — closes command injection via `remediation_id`
-- **trust-boundaries `isHighImpact`**: added chmod/chown/dd/mv/kill/shutdown/eval/command-substitution/redirect detection + shell-obfuscation normalization; narrowed interpreter-script pattern to stop false-positives on `node <projectfile>`
-- **eis-client**: `verifyRemoteProvenance` delegates to real ZTAI signature verification, fail-closed (was returning true for any non-empty signature)
-
-### Integrity / honest labeling (fixed)
-
-- **ztai-archiver**: `verifyIntegrity()` recomputes the Merkle root from the live log (was a no-op `return true`); fail-closed on tamper/delete/reorder
-- **mesh-self-healer**: emits an honest degraded advisory instead of fabricated 94%/100% consensus
-- **logic-validator**: probes Ollama when reachable, honest heuristic fallback; stopped advertising the dead model path
-- **reason-source-aligner**: consistent return shape (uninitialized no longer silently disables the mission-fidelity gate); real Jaccard similarity
-- **sre-manager**: HMAC artifact relabeled as integrity tag, not "ZK-Proof"
-- **installer-core**: PQAS-enabled message gated on the real `experimental.pqc_demo` flag
-- **finding-synthesizer**: detects real severity-gap contradictions (was hardcoded `[]`)
-- **logic-drift-detector**: relabeled heuristic, not "Neural"
-- **session-manager / shadow-mirror(docker) / regression-writer / skill-registry**: honest disclosure instead of silent empties / fake isolation / tautological tests / mock placeholders
-- **MINDFORGE.md**: `[PQAS_ENFORCED]` reconciled to reflect simulated/inactive default; **ztai-manager** logs relabeled `[ZTAI-HSM-SIM]`
-
-### CI / hygiene
-
-- **release workflow**: asserts the git tag matches `package.json` version (fail-closed) + skips publish if the version is already on npm
-- **version-check**: runtime drift check widened to all 4 sources (was 2); SDK README guarded by test
-- removed dead `AuditRotator` class (CHANGELOG had wrongly claimed it removed); deprecated orphaned `createAppendQueue`
-- refreshed stale `v11.1.0` banners → current; added `cost_routing.shadow_mode` latch
-- version bumped to 11.2.1 across all sources
-
-## [11.2.0] - 2026-05-31 — "Verification & Trust"
-
-### Added
-
-- **UC-08 — Unified Verification Runner**
-  - `bin/engine/verification-runner.js`: orchestrates test/lint/audit/typecheck stages into structured results with pass/fail/skip per stage
-  - `bin/engine/verify-cli.js`: CLI entrypoint writing `.planning/VERIFICATION.md` reports
-  - `mindforge verify` command registered in CLI
-
-- **UC-25 — Eval Harness**
-  - `bin/eval/eval-harness.js`: recall@k, nDCG (graded relevance), and `runEval()` orchestrator for measuring retrieval quality
-  - `bin/eval/golden-set-retrieval.json`: 10-query seed golden set covering orchestration, security, memory, cost, verification, hooks, and architecture domains
-
-- **UC-22 — Tool/MCP Trust Boundaries**
-  - `bin/security/trust-boundaries.js`: manifest pinning (deterministic SHA-256 with recursive key sort), tamper detection, untrusted output tagging with provenance, and high-impact command detection
-  - `bin/security/trust-gate-hook.js`: PreToolUse hook that blocks destructive Bash commands (rm -rf, force-push, DROP TABLE, hard reset) via native Claude Code hooks
-  - Trust-gate registered in `.claude/settings.json` PreToolUse
-
-- **Council CLI**
-  - `bin/council-cli.js`: thin wrapper wiring `runCouncil` to the `/mindforge:council` command with structured JSON output and formatted display
-
-### Fixed
-
-- `auto-runner.js`: removed erroneous `new` on singleton ZTAIManager instance (was line 692)
-
-### Removed
-
-- Dead `quantum-verify` CLI command entry (no handler existed)
-- Dead `AuditRotator` class and its export from `bin/utils/file-io.js` (zero callers; rotation broke the hash chain at rotation boundaries — see `bin/autonomous/audit-writer.js`)
-
-### Changed
-
-- Config `pqas_signing` provider clarified as "Dilithium-5 (simulated — inactive)"
-- Banner version strings updated to v11.1.0 in self-corrective-synthesizer and remediation-engine
-- SDK README heading updated to "New in v11.1.0"
-- Version bumped to 11.2.0 across package.json and config.json
-
----
-
-## [11.1.0] - 2026-05-31 — "Beast Mode"
-
-### Added
-
-- **Pillar I — Integrity & Trust**
-  - `bin/utils/append-queue.js`: single-writer fsync'd append queue for concurrent-safe durable writes (UC-09)
-  - `bin/governance/audit-hash.js`: single canonical SHA-256 hasher shared by writer + verifier (UC-04)
-  - `bin/governance/audit-verifier.js` + `bin/verify-audit.js` CLI: fail-closed hash-chain verifier (`exit 1` on any break) (UC-04)
-  - All 7 audit-write paths unified through `appendAuditEntrySync` — the audit log is now genuinely tamper-evident with a verifiable prev-hash chain (UC-04b)
-  - Simulated PQC demoted: `pqas_enabled=false` by default, gated behind `experimental.pqc_demo`; Tier-3 uses real Ed25519 (UC-24)
-
-- **Pillar II — Orchestration Correctness**
-  - `bin/autonomous/dependency-dag.js`: real Kahn topological sort + cycle detection, ported from test-only to the engine (UC-03)
-  - `planWaves(handoffs, {useDag:true})` opt-in DAG wave planning — explicit `.wave` always wins, legacy default unchanged (UC-03)
-  - Pre-flight cycle detection halts loud before any wave executes (UC-03)
-  - Wave-boundary timeout enforcement: `isTimedOut` + status `'timeout'` + resumable state (UC-14)
-  - Opt-in rollback hook records intent on terminal ESCALATE (no auto git-reset by default) (UC-14)
-  - `bin/engine/council-runtime.js`: minimal 4-voice council runtime (ADS loop) with injectable model, consensus scoring, dissent capture (UC-10)
-
-- **Pillar III — Cost-Aware Routing**
-  - `bin/models/pricing-registry.js`: single source of truth for model pricing, loaded from `config.json` market_registry (UC-05)
-  - All 3 providers routed through the registry — hardcoded pricing eliminated (UC-05)
-  - Prompt-cache accounting: `cache_control:{type:'ephemeral'}` on system blocks, `cache_read`/`cache_creation` parsed + priced at ~10% input rate (UC-21)
-  - `bin/models/difficulty-scorer.js`: heuristic 1-10 scorer with Tier-3 floor (UC-06)
-  - Shadow-mode difficulty routing: logs intended model without changing actual selection (UC-06)
-
-- **Pillar IV — Native Alignment + Observability**
-  - `.claude/settings.json` with hooks under real native events (`PostToolUse`/`PreToolUse`/`SessionStart`) — the security guard and context monitor now actually fire (UC-19a)
-  - `bin/hooks/instinct-capture-hook.js`: auto-captures behavioral patterns via PostToolUse, respects session limit (UC-11)
-  - `bin/engine/otel-exporter.js`: optional OTel GenAI exporter mapping NexusTracer spans to `gen_ai.*` semantic conventions, gated behind `OTEL_EXPORTER_OTLP_ENDPOINT` (UC-18)
-  - `bin/memory/retrieval-fusion.js`: Reciprocal Rank Fusion (RRF) over knowledge-graph + BM25 retrieval paths, replacing incomparable linear blends (UC-20)
-
-### Fixed
-
-- Audit-writer `close()` data-loss bug: threshold-triggered flushes were un-awaited, losing up to 10 entries on close (UC-09)
-- Audit-writer flush failure no longer crashes the process via unhandled rejection — logs to stderr (UC-09)
-- Vector-hub exit guard: `_dirty` boolean → pending-saves counter, closing a window where scheduled-but-unwritten saves were lost on hard exit (UC-09)
-- `isTimedOut` fails CLOSED on malformed `timeout_at` (garbage deadline = halt, not run unbounded) (UC-14)
-- Council position validation rejects NaN confidence / invalid recommendations (UC-10)
-- Dissent is now captured under NO_CONSENSUS verdict (the deadlock case where it matters most) (UC-10)
-- Pre-flight DAG check only cycle-checks stable-id tasks (id-less tasks can't be dependency targets) (UC-03)
-- `groupIntoWaves` self-defends: dangling dep throws "Unknown dependency" (not misleading "Circular") (UC-03)
-
-### Changed
-
-- Audit rotation retired (it broke the hash chain at every 5000-line boundary); AUDIT.jsonl grows unbounded for now — chain-aware compaction is a deferred future feature (UC-04b)
-- `knowledge-store.js` appends now use `appendDurableSync` (openSync+writeSync+fsyncSync+closeSync) for guaranteed durability (UC-09)
-- `auto-shadow.js` `generateShadowContext()` now fuses retrieval paths via RRF instead of a single-path linear blend (UC-20)
-
-### Removed
-
-- Orphaned `createAuditWriter` buffered path (zero production callers after UC-04b unification)
-- Hardcoded per-token pricing from `anthropic-provider.js`, `openai-provider.js`, `gemini-provider.js` (replaced by PricingRegistry)
-
-## [11.0.1] - 2026-05-30 — "Stability Patch"
-
-### Fixed
-
-- **Version drift**: reconciled `.mindforge/config.json` (was 10.7.0) with the 11.x line; added a fail-closed pre-flight version-consistency assertion (`bin/utils/version-check.js`) and a regression test that runs the migration and asserts its post-state, so the drift cannot silently return. (UC-01)
-- **Lint & dead code**: resolved all 109 ESLint errors, removed orphaned `bin/dashboard/team-tracker.js`, and made CI fail on any lint error. Fixed a latent `no-const-assign` runtime crash in `bin/review/ads-engine.js`. (UC-02)
-- **SDK `executeCommand` no-op**: replaced the published no-op stub (which made `batchExecute` report every task "fulfilled" while executing nothing) with a real `child_process.spawn` executor with stdout/stderr capture, timeout (SIGTERM→SIGKILL), and exit-code propagation; added a regression test against the compiled dist. (UC-07a)
-
-## [11.0.0] - 2026-05-28 — "Sovereign Stability"
-
-### Breaking Changes
-
-- `verifyZKProof()` returns structured `{ verified, reason }` instead of throwing
-- `signPQ()` returns `{ signature, simulated, algorithm }` object instead of raw string
-- Wave task execution order within waves is no longer deterministic
-- SDK bumped to 11.0.0 with new type exports
-- Dashboard tokens now expire after 24 hours
-- `TemporalHub.captureState()` and `rollbackTo()` are now async
-
-### Added
-
-- LRUMap utility class for bounded caches with eviction callbacks
-- Atomic JSON write primitives (write-to-temp, fsync, rename)
-- AUDIT.jsonl log rotation with gzip archival (max 5000 lines)
-- HANDOFF.json structural validation (fail-open)
-- Temporal snapshot garbage collection (retain 50, expire > 7 days)
-- BM25 scoring with document-length normalization
-- Persistent index cache (mtime-based invalidation)
-- Persistent adjacency index for knowledge graph
-- Correction effectiveness tracking in self-corrective synthesizer
-- Full remediation strategy implementations (CONTEXT_COMPRESSION, GOLDEN_TRACE_INJECTION, REASONING_RESTART)
-- Graduated intelligence interlock (+1/+2/MAX tier) with cost-awareness
-- 3-tier stuck detection (hash → length → truncated Levenshtein)
-- Adaptive context window (10/20/30 based on velocity)
-- Configurable external ZK verifier module path
-- Ephemeral SRE enclave keys (crypto.randomBytes)
-- Time-limited RBAC role elevation with auto-expiry
-- Session-scoped ZTAI agent registry
-- Dashboard rate limiting (100 req/min/IP) and token expiration (24h)
-- /api/v1/token/refresh endpoint
-- Optional GPG approval verification
-- GET /api/v1/system observability endpoint (heap, uptime, audit stats)
-- checkHeapHealth() with warning/critical thresholds
-- Remediation effectiveness persistence
-- Model router dynamic reload (mtime-based, 60s interval)
-- P95 latency ring buffer for cloud broker
-- EIS client with real fetch + 3-retry exponential backoff
-- Semaphore-based parallel wave execution (max concurrency configurable)
-- WebSocketEventStream with auto-reconnect
-- SDK streamExecution() with AsyncIterable<StreamChunk>
-- SDK batchExecute() with concurrent task execution
-- SDK validateRuntimeConfig()
-- Model streaming support (Anthropic, OpenAI, Gemini providers)
-- Migration script (bin/migrations/10.7.0-to-11.0.0.js)
-
-### Changed
-
-- sessionDriftHistory bounded to 500 entries via LRUMap
-- entropyCache bounded to 1000 entries via LRUMap
-- Cloud broker failure tracking uses 5-minute sliding window
-- Self-corrective synthesizer window expanded from 10 → 50 events
-- Context refactorer uses adaptive window instead of fixed 20
-
-### Fixed
-
-- Memory leaks from unbounded Maps in long-running sessions
-- Data corruption risk on process crash during state file writes
-- Disk exhaustion from unbounded AUDIT.jsonl and snapshot growth
-- Hardcoded SRE enclave private key (security issue)
-
----
-
-## [10.7.0] - 2026-05-27 — "Platform Sovereign"
-
-### Added (v10.7.0)
-
-- **10 new core skills** — internal-developer-platform, self-serve-infrastructure, platform-reliability, developer-productivity-metrics, api-marketplace, build-system-optimization, secrets-platform, environment-management, platform-observability, migration-platform.
-- **6 new commands** — `/mindforge:platform`, `/mindforge:build-opt`, `/mindforge:secrets-mgmt`, `/mindforge:environments`, `/mindforge:observability-platform`, `/mindforge:migration-mgmt`.
-- **6 new personas** — platform-lead, build-engineer, environment-engineer, productivity-analyst, secrets-engineer, migration-architect.
-- **1 new swarm template** — PlatformSwarmV2 (HITL platform engineering + migration).
-- **200 core skills milestone** — Full coverage across 12 domains.
-- **Swarm templates v15.0.0** — 49 total templates.
-
----
-
-## [10.6.0] - 2026-05-27 — "Data Alchemy"
-
-### Added (v10.6.0)
-
-- **10 new core skills** — causal-inference, feature-engineering, ml-monitoring, data-governance, stream-processing, data-lakehouse, experiment-platform, data-mesh, real-time-analytics, data-privacy-engineering.
-- **6 new commands** — `/mindforge:causal`, `/mindforge:lakehouse`, `/mindforge:data-mesh`, `/mindforge:stream`, `/mindforge:privacy-eng`, `/mindforge:realtime-analytics`.
-- **6 new personas** — causal-scientist, data-mesh-architect, stream-engineer, lakehouse-architect, privacy-engineer, analytics-engineer.
-- **1 new swarm template** — DataAlchemySwarm (HITL data architecture + privacy).
-
----
-
-## [10.5.0] - 2026-05-27 — "AI Frontier"
-
-### Added (v10.5.0)
-
-- **10 new core skills** — multimodal-ai, ai-safety-alignment, synthetic-data-generation, model-evaluation, embedding-systems, llm-orchestration, knowledge-graphs, ml-feature-store, ai-cost-management, autonomous-agents.
-- **8 new commands** — `/mindforge:multimodal`, `/mindforge:ai-safety`, `/mindforge:embeddings`, `/mindforge:llm-route`, `/mindforge:knowledge-graph`, `/mindforge:feature-store`, `/mindforge:ai-cost`, `/mindforge:agent-design`.
-- **8 new personas** — multimodal-engineer, ai-safety-engineer, embedding-architect, llm-orchestrator, knowledge-engineer, feature-store-engineer, ai-economist, agent-architect.
-- **1 new swarm template** — AIFrontierSwarm (HITL AI system architecture + safety).
-
----
-
-## [10.4.0] - 2026-05-27 — "Cross-Platform"
-
-### Added (v10.4.0)
-
-- **10 new core skills** — react-native-patterns, flutter-architecture, offline-first-design, progressive-web-app, mobile-performance, cross-platform-testing, app-store-deployment, mobile-security, responsive-native, push-notification-architecture.
-- **6 new commands** — `/mindforge:mobile`, `/mindforge:react-native`, `/mindforge:flutter`, `/mindforge:offline`, `/mindforge:pwa`, `/mindforge:push-notify`.
-- **6 new personas** — mobile-architect, react-native-engineer, flutter-engineer, offline-specialist, mobile-security-engineer, pwa-architect.
-- **1 new swarm template** — MobileSwarm (HITL cross-platform architecture).
-
----
-
-## [10.3.0] - 2026-05-27 — "Leader's Edge"
-
-### Added (v10.3.0)
-
-- **10 new core skills** — technical-leadership, mentoring-patterns, stakeholder-communication, conflict-resolution, incident-communication, hiring-engineering, delegation-patterns, meeting-architecture, performance-reviews, change-management.
-- **6 new commands** — `/mindforge:lead`, `/mindforge:communicate`, `/mindforge:hire`, `/mindforge:delegate`, `/mindforge:meeting-design`, `/mindforge:change`.
-- **6 new personas** — tech-lead-coach, communication-architect, hiring-strategist, change-agent, meeting-designer, mentorship-lead.
-- **1 new swarm template** — LeadershipSwarm (HITL engineering leadership).
-
----
-
-## [10.2.0] - 2026-05-27 — "Industry Forge"
-
-### Added (v10.2.0)
-
-- **10 new core skills** — healthcare-systems, fintech-patterns, ecommerce-architecture, gaming-backend, edtech-platform, saas-multi-tenant, media-streaming, iot-platform, marketplace-trust, logistics-optimization.
-- **8 new commands** — `/mindforge:healthcare`, `/mindforge:fintech`, `/mindforge:ecommerce`, `/mindforge:gaming`, `/mindforge:edtech`, `/mindforge:iot`, `/mindforge:marketplace`, `/mindforge:logistics`.
-- **8 new personas** — healthcare-engineer, fintech-architect, ecommerce-engineer, gaming-engineer, edtech-architect, iot-architect, marketplace-engineer, logistics-architect.
-- **1 new swarm template** — IndustryVerticalSwarm (HITL domain-specific architecture).
-- **150 core skills milestone** — Industry vertical coverage added.
-
----
-
-## [10.1.1] - 2026-05-26 — "Scale & Edge"
-
-### Added (v10.1.1)
-
-- **10 new core skills** — edge-computing, serverless-patterns, container-security, zero-trust-architecture, ai-agent-deployment, distributed-consensus, data-pipeline-design, dns-architecture, cdn-optimization, database-sharding-advanced.
-- **6 new commands** — `/mindforge:edge`, `/mindforge:serverless`, `/mindforge:zero-trust`, `/mindforge:agent-deploy`, `/mindforge:data-pipeline`, `/mindforge:cdn`.
-- **6 new personas** — edge-engineer, zero-trust-engineer, agent-ops-engineer, consensus-engineer, data-pipeline-architect, cdn-architect.
-- **2 new swarm templates** — EdgeScaleSwarm (HITL edge + CDN), DistributedSwarm (HITL consensus + pipelines).
-- **140 core skills milestone** — Comprehensive coverage of emerging technology and massive-scale patterns.
-- **Swarm templates v14.0.0** — Bump from v13.0.0 with 2 new templates (total: 43 swarm templates).
-
----
-
-## [10.1.0] - 2026-05-26 — "Strategic Intelligence"
-
-### Added (v10.1.0)
-
-- **20 new core skills** — build-vs-buy, technology-radar, architecture-tradeoff-analysis, technical-interview-design, post-incident-learning, team-topology-design, sprint-retrospective-facilitation, knowledge-sharing-systems, estimation-techniques, on-call-design, experiment-design, analytics-instrumentation, notification-system-design, payment-integration, email-deliverability, agent-memory-design, agent-evaluation-framework, multi-turn-conversation-design, agent-tool-selection, human-in-the-loop-design.
-- **10 new commands** — `/mindforge:build-vs-buy`, `/mindforge:tech-radar`, `/mindforge:team-topology`, `/mindforge:retro`, `/mindforge:experiment`, `/mindforge:analytics`, `/mindforge:payments`, `/mindforge:agent-memory`, `/mindforge:agent-eval`, `/mindforge:hitl`.
-- **8 new personas** — decision-architect, team-coach, knowledge-curator, experiment-designer, payments-engineer, agent-memory-designer, agent-evaluator, hitl-architect.
-- **3 new swarm templates** — DecisionSwarm (HITL decision quality), TeamDesignSwarm (HITL team topology), AgentMetaSwarm (autonomous self-improvement).
-- **130 core skills milestone** — Category expansion into decision science, team engineering, product patterns, and agent meta-intelligence.
-- **Minor version bump (10.1.0)** — Represents category expansion beyond pure engineering into strategy and meta-intelligence.
-- **Swarm templates v13.0.0** — Bump from v12.0.0 with 3 new templates (total: 41 swarm templates).
-
----
-
-## [10.0.9] - 2026-05-26 — "Full Spectrum"
-
-### Added (v10.0.9)
-
-- **20 new core skills** — streaming-architecture, queue-design, real-time-sync, cost-estimation, technical-debt-management, capacity-planning, graceful-degradation, idempotency-patterns, rate-limiting-design, code-generation-patterns, dependency-management, git-workflow-design, i18n-architecture, a11y-testing, multi-tenancy-patterns, audit-logging, database-performance, bundle-optimization, graphql-patterns, pagination-patterns.
-- **10 new commands** — `/mindforge:stream`, `/mindforge:queue`, `/mindforge:finops`, `/mindforge:tech-debt`, `/mindforge:degrade`, `/mindforge:idempotent`, `/mindforge:rate-limit`, `/mindforge:i18n`, `/mindforge:multi-tenant`, `/mindforge:graphql`.
-- **8 new personas** — streaming-engineer, finops-analyst, debt-manager, resilience-engineer, codegen-specialist, i18n-architect, multi-tenancy-architect, graphql-designer.
-- **3 new swarm templates** — StreamingSwarm (HITL real-time), ResilienceSwarm (autonomous failure engineering), GovernanceSwarm (HITL data governance).
-- **110 core skills milestone** — Comprehensive coverage across all major software engineering domains achieved.
-- **Swarm templates v12.0.0** — Bump from v11.0.0 with 3 new templates (total: 38 swarm templates).
-
----
-
-## [10.0.8] - 2026-05-26 — "Deep Patterns"
-
-### Added (v10.0.8)
-
-- **20 new core skills** — contract-testing, load-testing, mutation-testing, visual-regression-testing, monorepo-management, cli-design, developer-onboarding, error-handling-architecture, caching-strategies, migration-strategies, connection-pooling, event-driven-architecture, api-gateway-patterns, websocket-patterns, feature-flag-management, secrets-rotation, compliance-as-code, rag-architecture, fine-tuning-workflow, llm-cost-optimization.
-- **10 new commands** — `/mindforge:contract-test`, `/mindforge:load-test`, `/mindforge:monorepo`, `/mindforge:cli`, `/mindforge:cache`, `/mindforge:events`, `/mindforge:secrets`, `/mindforge:rag`, `/mindforge:feature-flags`, `/mindforge:compliance`.
-- **8 new personas** — contract-tester, dx-engineer, cache-architect, event-architect, compliance-engineer, ml-ops-engineer, platform-engineer, api-gateway-designer.
-- **3 new swarm templates** — TestingDeepSwarm (autonomous deep testing), PlatformSwarm (HITL platform engineering), MLOpsSwarm (HITL ML operations).
-- **90 core skills milestone** — Production-depth coverage for testing, caching, events, secrets, compliance, RAG, and cost optimization.
-- **Swarm templates v11.0.0** — Bump from v10.0.0 with 3 new templates (total: 35 swarm templates).
-
----
-
-## [10.0.7] - 2026-05-26 — "Meta Engineer"
-
-### Added (v10.0.7)
-
-- **20 new core skills** — prompt-engineering, context-engineering, agent-orchestration-patterns, tool-design, guardrails-and-safety, observability-stack, ci-cd-pipeline, infrastructure-as-code, incident-management, chaos-engineering, data-modeling, api-versioning, search-implementation, design-system, state-management, responsive-patterns, auth-patterns, supply-chain-security, technical-writing, code-review-methodology.
-- **10 new commands** — `/mindforge:prompt`, `/mindforge:context-budget`, `/mindforge:orchestrate`, `/mindforge:observability`, `/mindforge:pipeline`, `/mindforge:data-model`, `/mindforge:design-tokens`, `/mindforge:auth-flow`, `/mindforge:write-rfc`, `/mindforge:review-guide`.
-- **8 new personas** — prompt-architect, agent-orchestrator, sre-lead, pipeline-engineer, data-architect, design-system-lead, auth-engineer, technical-writer-lead.
-- **3 new swarm templates** — PromptEngineeringSwarm (HITL AI engineering), SRESwarm (HITL reliability), FrontendSwarm (autonomous design system).
-- **70 core skills milestone** — Framework now covers AI engineering, DevOps, reliability, data, frontend, advanced security, and technical communication.
-- **Swarm templates v10.0.0** — Bump from v9.0.0 with 3 new templates (total: 32 swarm templates).
-
----
-
-## [10.0.6] - 2026-05-26 — "Complete Arsenal"
-
-### Added (v10.0.6)
-
-- **17 new core skills** — microservices-patterns, cqrs-event-sourcing, system-design, business-analyst, product-manager, market-researcher, typescript-advanced, python-performance, react-performance, k8s-deployment, writing-plans, writing-skills, using-git-worktrees, code-tour, autonomous-agent-harness, mcp-server-patterns, proofreader.
-- **10 new commands** — `/mindforge:microservices`, `/mindforge:system-design`, `/mindforge:brd`, `/mindforge:product-spec`, `/mindforge:market-research`, `/mindforge:code-tour`, `/mindforge:mcp-server`, `/mindforge:proofread`, `/mindforge:worktrees`, `/mindforge:plan-write`.
-- **8 new personas** — business-analyst, product-owner, market-analyst, mcp-designer, proofreader, system-designer, worktree-manager, code-narrator.
-- **3 new swarm templates** — ArchDesignSwarm (HITL system design), ProductSwarm (HITL product strategy), DocumentationSwarm (autonomous content quality).
-- **Swarm templates v9.0.0** — Bump from v8.0.0 with 3 new templates (total: 29 swarm templates).
-- **50 core skills milestone** — Framework now covers architecture, business, languages, workflow, infrastructure, and documentation domains.
-
----
-
-## [10.0.5] - 2026-05-26 — "Forge Master"
-
-### Added (v10.0.5)
-
-- **5 new core skills** — skill-creator-meta, deployment-workflow, dmux-workflows, vibe-security, instinct-clustering.
-- **5 new commands** — `/mindforge:create-skill`, `/mindforge:deploy`, `/mindforge:dmux`, `/mindforge:vibe-check`, `/mindforge:cluster-instincts`.
-- **5 new personas** — skill-smith, deployment-captain, dmux-orchestrator, vibe-checker, saga-orchestrator.
-- **2 new swarm templates** — DeploymentSwarm (HITL staged rollout), ForgeSwarm (autonomous skill creation).
-- **De-slop gate** — Phase 6.5 in verification-loop: informational de-slop scan before shipping (non-blocking).
-- **Cross-model eval spec** — `.mindforge/engine/cross-model-eval.md` for routing same task to 2 models and comparing outputs.
-- **Swarm templates v8.0.0** — Bump from v7.0.0 with 2 new templates (total: 26 swarm templates).
-
----
-
-## [10.0.4] - 2026-05-26 — "Santa's Eval"
-
-### Added (v10.0.4)
-
-- **8 new core skills** — santa-method, eval-harness, quality-audit, testing-anti-patterns, defense-in-depth, codebase-onboarding, rfc-pipeline, de-sloppify.
-- **6 new commands** — `/mindforge:santa`, `/mindforge:eval`, `/mindforge:quality-audit`, `/mindforge:rfc`, `/mindforge:onboard`, `/mindforge:de-slop`.
-- **6 new personas** — eval-judge, rfc-architect, anti-pattern-hunter, onboarding-navigator, de-sloppifier, quality-scorer.
-- **3 new swarm templates** — EvalSwarm (autonomous eval gate), OnboardingSwarm (autonomous codebase discovery), RFCSwarm (HITL spec decomposition).
-- **Proactive Skill Suggestion Engine** — Signal-based skill detection (file/error/task patterns) with confidence threshold (0.7), cooldown tracking, and debounce logic.
-- **Eval storage** — `.mindforge/evals/` directory for persisting eval configs, rubrics, and results.
-- **Swarm templates v7.0.0** — Bump from v6.0.0 with 3 new templates (total: 24 swarm templates).
-
----
-
-## [10.0.3] - 2026-05-25 — "Council Awakens"
-
-### Added (v10.0.3)
-
-- **10 new core skills** — agent-loops, multi-llm-consult, continuous-learning, council, verification-loop, threat-modeling, autonomous-loops, agent-introspection-debugging, cost-aware-routing, doc-health-audit.
-- **8 new commands** — `/mindforge:council`, `/mindforge:consult`, `/mindforge:verify-loop`, `/mindforge:introspect`, `/mindforge:cost-report`, `/mindforge:threat-model`, `/mindforge:learn-instinct`, `/mindforge:evolve-skills`.
-- **9 new personas** — cost-optimizer, threat-modeler, council-architect, council-skeptic, council-pragmatist, council-critic, instinct-curator, doc-auditor, multi-model-bridge.
-- **3 new swarm templates** — CouncilSwarm (HITL decision gate), VerificationSwarm (autonomous quality gates), LearningSwarm (instinct management).
-- **Skill Composition System** — Skills can now declare dependencies on other skills via `compose:` frontmatter field. Composed skills are injected as summaries (max 2-level depth, cycle detection).
-- **Instinct Engine** — Auto-capture learned behaviors with confidence scoring. Instincts auto-promote to skills at 0.85 confidence after 5+ successful applications. Project-scoped, max 100 per project.
-- **Cost Tracking Module** — Token budgeting, 5-tier model routing (Haiku/Sonnet/Opus/Gemini/GPT-4o), spend analytics via token-ledger.jsonl.
-- **Council Framework** — 4-voice decision harness (Architect, Skeptic, Pragmatist, Critic) with weighted consensus scoring, dissent documentation, and 5 pre-built templates.
-- **Cross-Iteration Bridge** — SHARED_TASK_NOTES.md for semantic context persistence across autonomous mode iterations (complements HANDOFF.json).
-- **Swarm templates v6.0.0** — Bump from v5.0.0 with 3 new templates (total: 21 swarm templates).
-- **claude-opus-4-7** added to market_registry in config.json.
-- **Loader composition step** — New Step 4.1 in skill loader for resolving `compose:` dependencies.
-
----
-
-## [10.0.2] - 2026-05-24 — "Persona Expansion"
-
-### Added (v10.0.2)
-
-- **47 new specialist personas** — Security & Compliance (6), Architecture & System Design (8), Frontend & UX (10), Performance & Reliability (5), Data & ML (4), DevOps & Infrastructure (6), Language Specialists (5), DX & Operations (8), Strategy & Review (9), Specialized Engineering (10).
-- **6 new swarm templates** — ArchitectureSwarm, PerformanceSwarm, InfrastructureSwarm, AccessibilitySwarm, ReviewSwarm, MigrationSwarm.
-- **Updated existing swarms** — UISwarm, BackendSwarm, SecuritySwarm, DeveloperExperienceSwarm, DataMeshSwarm, IncidentResponseSwarm, ComplianceSwarm, QualityAssuranceSwarm now include relevant new specialist personas.
-- **Swarm templates v5.0.0** — Bump from 4.2.0 with expanded member rosters and new specialist orchestration patterns.
-- **Persona registry documentation** — 10 new category tables in `docs/registry/PERSONAS.md`.
-- **Persona reference guide** — Updated `docs/PERSONAS.md` with all 47 entries (total: 108 personas).
-
----
-
-## [10.0.0] - 2026-05-21 — "Bedrock Fortified"
-
-### Added (v10.0.0)
-
-- **Bearer token authentication** — All dashboard mutating endpoints now require `Authorization: Bearer <token>` header. Protects SSE, steering, and approval routes.
-- **Auth on browser daemon** — The `/evaluate` endpoint on the Playwright browser daemon now requires authentication, closing a remote code execution vector.
-- **Unified test runner** — `tests/run-all.js` discovers and executes all 43 test files with pass/fail summary and exit code propagation.
-- **Shared utilities layer** — `bin/utils/paths.js`, `file-io.js`, and `errors.js` provide consistent path resolution, safe file I/O, and typed error classes across the codebase.
-- **AuditWriter module** — Buffered async writes with Merkle hash chaining, replacing direct synchronous file append calls in the auto-runner.
-- **State manager module** — Extracted from auto-runner for clean separation of execution state persistence.
-- **Task dispatcher module** — Extracted from auto-runner for isolated task routing and repair-operator logic.
-- **Wave executor module** — Extracted from auto-runner for wave-level orchestration and progress tracking.
-- **Dependabot configuration** — Weekly npm dependency updates and monthly GitHub Actions version bumps.
-- **CODEOWNERS file** — Defines review ownership for security-sensitive paths (`bin/governance/`, `bin/engine/`, SDK).
-- **npm provenance** — Release workflow now publishes with `--provenance` for SLSA Build Level 2 supply chain attestation.
-- **CLI "Did you mean?" suggestions** — Levenshtein distance matching for mistyped commands (e.g., `mindforge statsus` suggests `status`).
-- **CLI --verbose flag** — Enables detailed execution tracing for debugging framework behavior.
-- **CLI mindforge bin entry** — Post-install accessible binary for direct invocation without `npx`.
-- **Structured action allowlist** — Replaces the regex-based prompt injection blocklist with an explicit allowlist of permitted autonomous actions.
-- **.npmignore** — Excludes runtime artifacts, test fixtures, planning state, and intelligence logs from the published package.
-- **Knowledge store O(1) index** — Hash-map index for stores exceeding 100 entries, eliminating linear scans.
-- **SSE idle detection** — Dashboard polling pauses when no clients are connected, reducing CPU and I/O overhead.
-- **Metrics TTL cache** — 5-second time-to-live cache prevents redundant metrics recomputation on rapid dashboard refreshes.
-- **Smart mtime-based polling** — File watchers skip re-reads when mtime has not changed since last poll cycle.
-- **NON-OVERRIDABLE governance** — Added non-overridable parameter section to MINDFORGE.md for security-critical settings.
-- **HANDOFF schema alignment** — Unified HANDOFF.json to v1.0.0 schema with all required fields.
-
-### Changed (v10.0.0)
-
-- **VectorHub: sql.js (WASM) replaces better-sqlite3 (native C++)** — Eliminates native compilation, node-gyp, and platform-specific build failures. Zero native dependencies.
-- **VectorHub: lazy Proxy + factory pattern** — Database connection is deferred until first query. Callers use `createVectorHub()` factory instead of direct constructor.
-- **VectorHub: WAL mode + FTS4** — Write-Ahead Logging for concurrent reads; FTS4 replaces FTS5 for sql.js compatibility.
-- **VectorHub: parameterized queries throughout** — All user-influenced values use bound parameters, closing SQL injection vectors.
-- **Auto-runner decomposition** — Reduced from 672 lines to 367 lines by extracting 4 focused modules (task-dispatcher, wave-executor, state-manager, audit-writer).
-- **NexusTracer async migration** — Moved from synchronous file appends to non-blocking writes with back-pressure handling.
-- **PolicyEngine async migration** — Governance evaluation no longer blocks the event loop during high-throughput wave execution.
-- **SDK memory.ts rewrite** — Now self-contained with no imports from `../../bin/`. Independently publishable.
-- **SDK exports field** — Package.json `exports` map enables tree-shaking and prevents deep import of internal modules.
-- **SDK files field** — Only ships compiled output and type declarations; source maps excluded from production.
-- **Real validateConfig() implementation** — Validates 5 required fields with type checking and schema enforcement (replaces no-op stub).
-- **Lazy module loading in auto-runner** — 12 heavy `require()` calls deferred to first-use, cutting cold-start time.
-- **Package size reduction** — From 3.9 MB / 866 files to 1.1 MB / 245 files (72% smaller).
-- **CI workflow** — Node 18/20/22 matrix, `npm ci`, npm caching, c8 coverage at 30% threshold.
-- **Model topology assertions** — Updated test expectations to match Claude 4.x and Gemini 2.5 model family.
-- **Skills MANIFEST** — Updated manifest to register all 10 core skills with proper paths and schema version.
-- **README rewritten** — Clear value proposition, quick-start, and architecture overview at the top.
-- **All docs version-stamped** — References updated to v10.0.0 throughout.
-- **SDK reference documentation** — Expanded from 28 lines to 230+ lines with full API surface coverage.
-
-### Fixed (v10.0.0)
-
-- **CRITICAL: Hardcoded npm token removed** — `NPM_TOKEN` was embedded in CI config; now sourced exclusively from environment variables with startup validation.
-- **CRITICAL: ZK-proof governance bypass closed** — `verifyZKProof()` previously returned `true` unconditionally; now throws `GovernanceViolationError` with fail-closed semantics.
-- **CRITICAL: Path traversal in temporal API** — `auditId` parameter accepted `../` sequences; now validated with regex allowlist and directory containment check.
-- **CRITICAL: Command injection in change-classifier** — Shell-interpreted user input replaced with argument-array invocation, eliminating shell interpretation.
-- **Runtime crash in learning-manager.js** — Null reference on empty knowledge store; added defensive initialization.
-- **4 broken test files** — Fixed `nexus-tracing.test.js`, `model-broker.test.js`, `feedback-loop.test.js`, and `security-audit.test.js` (stale mocks, missing fixtures, async race conditions).
-- **Observability workflow trigger** — Workflow `name:` field now matches the `workflow_run.workflows` reference in dependent jobs.
-- **Duplicate release-plane.yml** — Removed conflicting workflow file that caused CI confusion.
-
-### Removed (v10.0.0)
-
-- **better-sqlite3 dependency** — Native C++ addon removed in favor of sql.js (pure WASM). Eliminates `node-gyp`, Python, and platform-specific build toolchains.
-- **Dead CLI routes** — `sync-jira` and `sync-confluence` commands removed (integration stubs with no implementation).
-- **Duplicate release workflow** — `release-plane.yml` deleted; single `release.yml` is the canonical publish pipeline.
-- **Regex-based prompt injection blocklist** — Replaced by structured action allowlist (fewer false positives, auditable).
-
----
-
-## [9.0.0] - 2026-04-30
-
-### Added (v9.0.0: Bedrock Meridian — Pillars XXIV-XXVIII)
-
-- **Pillar XXIV: Grounded Wave Execution** — Replaced AutoRunner stubs with real HANDOFF-driven wave parsing, task dispatch, progress tracking, and state persistence. `hasNextWave()` reads HANDOFF.json wave groups. `executeWave()` iterates tasks with audit logging and repair-operator integration. `runPreFlight()` validates project state and restores progress from `auto-state.json`.
-- **Pillar XXV: Model Topology Modernization** — Updated all model references from Claude 3.x to Claude 4.x family (`claude-opus-4-7`, `claude-sonnet-4-6`, `claude-haiku-4-5`). Updated `model-router.js`, `model-client.js`, `model-broker.js`, `cloud-broker.js`, and `MINDFORGE.md`. Aligned pricing tables and fallback chains.
-- **Pillar XXVI: Unified Memory Architecture** — Added `knowledge` and `graph_edges` tables to VectorHub (SQLite). New `saveKnowledge()`, `searchKnowledge()`, `saveEdge()`, `getEdges()` methods. FTS5 index on knowledge content. Consolidates 4 JSONL-based memory systems into a single SQLite store.
-- **Pillar XXVII: Schema Migration Engine** — Added `_migrations` table to VectorHub for tracking applied migrations. New `v9-unified-memory.js` migration script that reads legacy JSONL stores and imports into SQLite. Registered in `migrate.js` runner.
-- **Pillar XXVIII: Integration Test Chain** — New `tests/v9-integration-chain.test.js` verifying the full pipeline: wave parsing, model topology, VectorHub schema, migration engine, and SDK sync. 17 assertions across all v9 pillars.
-
-### Changed (v9.0.0)
-
-- **SDK**: Bumped `@mindforge/sdk` VERSION to `9.0.0`. Added `WaveExecutionResult` and `MigrationResult` types. Added `readAutoState()` and `isDatabaseInitialized()` to client.
-- **VectorHub**: Removed try/catch `ALTER TABLE` pattern (v8 legacy). Schema creation is now declarative via `createTable().ifNotExists()`.
-- **MINDFORGE.md**: Version bumped to `9.0.0-BEDROCK`. Model topology section normalized to Claude 4.x.
-
----
-
-## [8.2.1] - 2026-04-25
-
-### Fixed (v8.2.1: Stability & Cleanup)
-
-- **Payload Pruning**: Removed development-only test scripts from `bin/engine` and `bin/governance`.
-- **State Sanitization**: Cleaned up local persistence files (`celestial.db`, `memory/*.jsonl`) and session-specific planning artifacts.
-- **Structural Hardening**: Restored clean templates for `.planning/HANDOFF.json` and other core descriptors.
-- **NPM Optimization**: Reduced package size by excluding unneeded build-time caches and logs.
-
-## [8.2.0] - 2026-04-18
-
-### Added (v8.2.0: Autonomous SRE Layer — Pillars XX-XXIII)
-
-- **Sentinel Engine**: Implemented `bin/sre/sentinel.js` for continuous observability and anomaly detection in the audit stream.
-- **Shadow Mirror Isolation**: Hybrid replication system supporting Level 1 (Git) and Level 2 (Docker) deterministic isolation.
-- **Adversarial SRE Debate**: Three-way consensus protocol for high-fidelity remediation auditing.
-- **SLI-Gating Protocol**: Metric-driven verification loops to prevent performance regressions during self-healing.
-- **Persona Hardening**: Registered `sre-engineer` and `sre-auditor`, with the Auditor locked to **Claude 4.5 Opus**.
-
-## [8.1.0] - 2026-04-13
-
-### Added (v8.1.0: Sovereign Identity — Pillar XIX)
-
-- **Sovereign Identity Synthesis**: Autonomous creation and evolution of `SOUL.md` from execution traces.
-- **Identity Hegemony**: Promoted `SOUL.md` to #1 Source of Truth in the instruction hierarchy.
-- **XML-Structured Workflows**: Standardized project initialization and identity management with high-fidelity XML prompting.
-- **Identity Engine**: Implemented `bin/memory/identity-synthesizer.js` with v8.1 mirroring heuristics.
-
----
-
-## [8.0.0] - 2026-04-12
-
-### Added (v8.0.0: Celestial Orchestration - SQLite & FMS)
-
-- **Pillar XV: Unified Persistence Layer**: Transitioned from file-based JSONL state to a high-performance **SQLite** engine (`celestial.db`) with **FTS5** semantic reasoning support.
-- **Pillar XVI: Federated Mesh Synthesis (FMS)**: Inter-project knowledge sharing via signed **MindForge Bundles (`.mfb`)** using ZTAI-signed identities.
-- **Pillar XVII: Autonomous Skill Evolution (ASE)**: Implemented `SkillEvolver` to mine "Golden Traces" and synthesize reusable agentic skills.
-- **Pillar XVIII: Orbital Governance**: Hardware-bound (HSM/Biometric) attestation for catastrophic-risk mutations (>95).
-- **Core Vector Expansion**: Upgraded `VectorHub` with `better-sqlite3` and `Kysely` for type-safe, sub-millisecond persistence.
-
----
-
-## [7.0.0] - 2026-04-11
-
-### Added (v7.0.0: Sovereign Intelligence - Post-Quantum & Homing)
-
-- **Pillar IX: Autonomous Resource Harvesting (ARH)**: Real-time token arbitrage and dynamic task-to-model steering based on MIR.
-- **Pillar X: Neural Drift Remediation (NDR)**: Logic-drift detection and automated reasoning recovery via semantic density heuristics.
-- **Pillar XI: Post-Quantum Agentic Security (PQAS)**: Lattice-based (Dilithium-5) signatures and ZK-Proof bypasses for future-proof identity.
-- **Pillar XII: Proactive Semantic Homing**: Autonomous "Intent Hunting" where agents proactively claim tasks from the Federated Intelligence Mesh (FIM).
-
----
-
-## [6.7.0] - 2026-04-09
-
-### Added (v6.7.0: Federated Intelligence Mesh - FIM Expansion)
-
-- **Architectural Health Tracking**: Aggregates RSA and IDC scores into collective project health metrics via `PillarHealthTracker`.
-- **Stability Pattern Extraction**: Captures high-efficacy SCS homing instructions as durable, sharable knowledge entries (`stability_pattern`).
-- **AutoRunner Lifecycle Hooks**: Automated capture at phase completion for organizational sync.
-
----
-
-## [6.6.0] - 2026-04-08
-
-### Added (v6.6.0: Self-Corrective Synthesis - SCS)
-
-- **Autonomous Self-Healing**: Integrated `SelfCorrectiveSynthesizer` into the `AutoRunner` wave loop.
-- **Reasoning Drift Recovery**: Triggered when Mission Fidelity (RSA) drops below 0.50, synthesizing a "Homing Instruction" anchored to specific requirements.
-- **Homing Signal Injection**: Automatically injects high-density refocus signals into the wave context to recover mission fidelity.
-
----
-
-## [6.5.0] - 2026-04-08
-
-### Added (v6.5.0: Reason-Source Alignment - RSA)
-
-- **Requirement-to-Thought Mapping**: Integrated `ReasonSourceAligner` into the core Auto-Runner loop.
-- **Mission Fidelity Tracking**: Every reasoning thought is now cross-referenced with `REQUIREMENTS.md` and tagged with a best-match REQ-ID.
-- **Mission Drift Detection**: Real-time monitoring of "Mission Alignment" confidence scores to prevent autonomous side-questing.
-- **Requirement Schema**: Established formal `[REQ-ID]` Markdown schema for project specs.
-
----
-
-## [6.4.0] - 2026-04-08
-
-### Added (v6.4.0: Context Entropy Guard - CEG)
-
-- **Context Entropy Guard (CEG)**:
-  - Implemented `context-entropy-guard.js` for active noise suppression and context pruning.
-  - Added semantic compression to `HandoverManager`, automatically summarizing stagnant reasoning loops into digests.
-  - Hardened Nexus Bundles against "Reasoning Entropy" (similarity-based token bloat).
-
----
-
-## [6.3.0] - 2026-04-08
-
-### Added (v6.3.0: Intelligence-Drift Coupling - IDC)
-
-- **Dynamic Model Upgrading (IDC)**:
-  - Implemented `intelligence-interlock.js` to bridge NDR (Neural Drift Remediation) and ARH (Resource Arbitrage).
-  - Added "Critical Drift" alert thresholds to `LogicDriftDetector`.
-  - Integrated IDC signals into `AutoRunner`, allowing for proactive model tier upgrades when reasoning decay is detected.
-  - Hardened `RouterSteering` to respect mid-wave MIR overrides.
-
----
-
-## [6.2.0] - 2026-04-02
-
-### Added (v6.2.0: Sovereign Enterprise Stable)
-
-- **Autonomous Agentic Learning (Double-Hook)**:
-  - Implemented `AGENTS_LEARNING.md` as a persistent, project-level engineering memory.
-  - Added automated hooks to capture architectural patterns and anti-patterns during phase execution.
-
-- **Enterprise-Grade Professionalization**:
-  - Hardened and professionalized the entire documentation suite (Architecture, Security, Universal Commands).
-  - Hardened framework branding across all system identifiers and logs.
-  - Updated agent registry and workflow atlas for v6 consistency.
-
-- **Identity Layer Hardening**:
-  - Professionalized source code logs and identifiers for enterprise-compliance.
-  - Test suite branding and identity alignment.
-
----
-
-## [6.2.0-alpha] - 2026-03-29
-
-### Added (v6.2.0-alpha: Sovereign Pillars XI & XII)
-
-- **Pillar XI: Post-Quantum Agentic Security (PQAS)**:
-    - Implemented `QuantumCrypto` for simulated lattice-based (Dilithium-5) signatures and ZK-Proof generation.
-    - Integrated `QuantumSafeKeyProvider` into `ZTAIManager` for Tier 4 (Quantum-Safe) agent identities.
-    - Hardened `PolicyEngine` with dynamic Biometric Challenges for Risk > 95 and ZK-Proof verification for high-leverage bypasses.
-- **Pillar XII: Proactive Semantic Homing**:
-    - Implemented `IntentHarvester` for proactive, idle-state task claiming from the Federated Intelligence Mesh (FIM).
-    - Implemented `MeshSelfHealer` for peer-to-peer collaborative reasoning and recovery from logic drift > 80.
-    - Integrated proactive behaviors into the `AutoRunner` core loop.
-- **Hardening & Verification**:
-    - Constant-time signature verification and cycle-detection for mesh healing.
-    - 100% pass rate on v7 specialized security and homing integration test suites.
-
-## Top Summary (v6.2.0-alpha)
-
-MindForge v6.2.0-alpha achieves full **Sovereign Intelligence** status by hardening the mesh against quantum-era threats and transitioning from reactive wave-processing to proactive agentic homing.
-
-## Highlights (v6.2.0-alpha)
-
-- **Quantum-Safe Identity (PQAS)**: Verifiable agentic signatures resistant to future cryptographic threats.
-*   **Proactive Homing**: Autonomous "Intent Hunting" that eliminates idle latency in the agentic mesh.
-*   **Edge-Case Biometrics**: Hardware-locked governance for catastrophic-risk operations (>95 Blast Radius).
-
----
-
-## [6.1.0-alpha] - 2026-03-29
-
-### Added (v6.1.0-alpha: Sovereign Intelligence)
-
-- **Pillar IX: Autonomous Resource Harvesting (ARH)**: 
-    - Implemented `MarketEvaluator` for real-time token arbitrage.
-    - Dynamic task-to-model steering based on MIR (Min-Intelligence-Requirement).
-    - ROI Telemetry integration for tracking agentic cost-savings.
-- **Pillar X: Neural Drift Remediation (NDR)**:
-    - Implemented `LogicDriftDetector` with semantic density and repetition heuristics.
-    - Integrated `RemediationEngine` for automated reasoning recovery (Injection/Restart).
-    - Hardened `NexusTracer` hooks for real-time drift sensing.
-- **Hardened Sovereign Heuristics**: Improved repetitive logic detection using punctuation-aware token analysis.
-- **ROI Arbitrage Ledger**: New audit event `roi_arbitrage_event` in `AUDIT.jsonl`.
-
-## Top Summary (v6.1.0-alpha)
-
-MindForge v6.1.0-alpha completes the **Sovereign Intelligence** phase. The framework now possesses the ability to quantify its own economic value through resource harvesting and self-heal from cognitive decay through neural drift remediation.
-
-## Highlights (v6.1.0-alpha)
-
-- **Autonomous FinOps (ARH)**: Real-time steering between tiered models based on task MIR.
-- **Self-Healing Reasoning (NDR)**: Proactive interruption of logic loops and hallucinations.
-- **Unified Integration**: Seamless integration of NDR/ARH into the core `NexusTracer` loop.
-
----
-
-## [6.0.0-alpha] - 2026-03-29
-
-The v6.0.0-alpha release introduces the final pillar of the MindForge Enterprise architecture: AgRevOps (Agentic Revenue Operations). This engine provides real-time ROI attribution for autonomous waves, enabling enterprises to quantify the business value of every agentic reasoning cycle.
-
-## Highlights (v6.0.0-alpha)
-
-- **AgRevOps ROI Engine**: Real-time value attribution for autonomous task execution.
-- **ROI Telemetry**: New `roi-telemetry.jsonl` stream for tracking cost efficiency vs. outcome quality.
-- **Nexus Steering Sync**: Final synchronization of the `NexusTracer` and `NexusSteering` protocols into the core execution loop.
-- **Protocol Automation**: Automated activation of `_extended` mindforge skills via the Neural Orchestrator.
-
----
-
-# MindForge v5.9.0 — Enterprise-Grade Hardening (Nexus Unification)
-
-## Top Summary (v5.9.0)
-
-The v5.9.0 release elevates the MindForge Enterprise architecture to "Enterprise-Grade" by unifying the tracing infrastructure into a single, high-fidelity asynchronous ART protocol and hardening the governance and arbitrage pillars with advanced cryptographic and resilience patterns.
-
-## Highlights (v5.9.0)
-
-- **Unified NexusTracer Singleton**: Migration of all tracing and reasoning capture to `bin/engine/nexus-tracer.js`. Standardized as a singleton with mandatory `async` methods to support ZTAI cryptographic signing.
-- **Merkle-Style Audit Integrity**: Hardened SRE and ZTAI logs with Merkle-root cumulative hash chains, ensuring every audit entry is cryptographically linked to the entire session history.
-- **MCA Circuit Breakers**: Stateful provider blacklisting in `CloudBroker` that automatically disables failing models for 10 minutes after 3 consecutive errors.
-- **Intelligence Metrics Decay**: Historical performance data now favors recent trends via a 0.95 decay factor, ensuring routing adaptivity.
-- **Async Test Suite**: 100% restoration of the core test suite (Nexus, SRE, RES) for the new asynchronous execution model.
-
----
-
-# MindForge v5.8.0 — Sovereign Reason Enclaves (ZK-Audit)
-
-## Top Summary (v5.8.0)
-
-The v5.8.0 release implements the sixth pillar of the Hyper-Enterprise roadmap: Sovereign Reason Enclaves with Zero-Knowledge (ZK) Audit Trails. This enables agents to provide cryptographic proof of policy adherence for confidential reasoning without exposing proprietary content to global logs.
-
-## Highlights (v5.8.0)
-
-- **ZK-Proof Compliance Certificates**: Simulated zero-knowledge proofs (DID-signed) for SRE sessions.
-- **Privacy-Preserving Auditing**: `NexusTracer` replaces raw thought traces with verifiable certificates for isolated tasks.
-- **Enclave Verification**: New `verifyZKProof` utility for non-custodial audit integrity checks.
-- Synchronized `mindforge-cc` documentation suite for Pillar VII & VIII.
-- Updated `CLAUDE.md` with standard `_extended` protocol awareness.
-
-## [5.9.0] - 2026-03-28
-
-### Added (v5.9.0)
-
-- **Pillar VII: Nexus Steering (Hierarchical Intent Orchestration)**.
-- Implemented `NexusTracer` for multi-layered reasoning visualization.
-- Added `NexusSteering` protocol for cross-session intent persistence.
-- Hardened `SwarmController` with `context7-depth` logic for high-complexity tasks.
-
-## [5.8.0] - 2026-03-28
-
-### Added (v5.8.0)
-
-- **Pillar VI: Sovereign Reason Enclaves (ZK-Proof Audit Trails)**.
-- Implemented `SRE-ISOLATED` reasoning mode in `SREManager`.
-- Generated **ZK-Proof Compliance Certificates** (signed by System DID) for confidential reasoning cycles.
-- Integrated masked audit logging in `NexusTracer`, replacing raw thought traces with verifiable proofs.
-- Added `verifyZKProof` utility for non-custodial audit verification.
-
-## [5.7.0] - 2026-03-28
-
-### Added (v5.7.0)
-
-- **Pillar V: Multi-Cloud Arbitrage (Task-to-Model Affinity Routing)**.
-- Implemented **Performance-Based Affinity Matrices** in `CloudBroker` for intelligence-first routing.
-- Added `performance-stats.json` persistence to track cross-provider success rates by task taxonomy.
-- Integrated automated result recording in `WaveFeedbackLoop`.
-- Prioritized **Probability of Success** over raw cost/latency in routing scoring.
-
-## [5.6.0] - 2026-03-28
-
-### Added (v5.6.0)
-
-- **Pillar IV: Supply Chain Trust (Binary Runtime Attestation)**.
-- Cryptographic skill signing in `SkillRegistry` via ZTAIManager Tier 3 Enclaves.
-- JIT Attestation in `SkillValidator` to verify skill integrity before agent execution.
-- `SIGNATURES.json` tracking for all enterprise-grade skills.
-
-## [5.5.0] - 2026-03-28
-
-### Added (v5.5.0)
-
-- **Pillar III: Predictive Agentic Reliability (Reasoning Entropy Monitoring)**.
-- Reasoning Entropy Scoring (RES) in `NexusTracer` to detect semantic stagnation and loops.
-- Proactive Self-Healing trigger for high-similarity thought sequences.
-- Steering Vector generation in `TemporalHindsight` to break agentic deadlocks.
-
-## [5.4.0] — Enterprise-Grade Hardening — 2026-03-28
-
-# Release Notes - MindForge v5.4.0 — Enterprise Resilience (Hardened Edition)
-
-This update elevates the v5.3.0 "Hyper-Enterprise" features to maximum robustness ("Production-Grade"), implementing critical safety systems and advanced observability.
-
-### Added
-
-- **Circuit Breaker Pattern**: Implemented a stateful `CircuitBreaker` in `federated-sync.js` to prevent network floods. Automatically disables mesh sync for 1 hour after 3 consecutive EIS failures.
-- **Critical-Path Protection**: Automated "Blast Radius" score of 100 in `impact-analyzer.js` for sensitive files (`.env`, `*.pem`, `id_rsa`, `package-lock.json`, etc.).
-- **Recursive Depth Penalty**: Introduced a 1.5x impact multiplier for actions deeper than 5 directory levels, preventing mass-scale silent modifications.
-- **Failure Telemetry**: Added `sync-history.jsonl` and `sync-telemetry.jsonl` for detailed conflict resolution and error auditability.
-- **Resilient Execution**: Raised default scores for `EXECUTE` and `GRANT` actions to increase governance oversight.
-
-🚀 **MindForge v5.4.0 — Enterprise Resilience (Hardened Edition)**
-
-This update elevates the v5.3.0 "Hyper-Enterprise" features to maximum robustness ("Production-Grade"), implementing critical safety systems and advanced observability.
-
-### 🛡️ Enterprise-Grade Hardening (v5.4.0)
-
-- **Circuit Breaker Pattern**: Implemented a stateful `CircuitBreaker` in `federated-sync.js` to prevent network floods. Automatically disables mesh sync for 1 hour after 3 consecutive EIS failures.
-- **Critical-Path Protection**: Automated "Blast Radius" score of 100 in `impact-analyzer.js` for sensitive files (`.env`, `*.pem`, `id_rsa`, `package-lock.json`, etc.).
-- **Recursive Depth Penalty**: Introduced a 1.5x impact multiplier for actions deeper than 5 directory levels, preventing mass-scale silent modifications.
-- **Failure Telemetry**: Added `sync-history.jsonl` and `sync-telemetry.jsonl` for detailed conflict resolution and error auditability.
-- **Resilient Execution**: Raised default scores for `EXECUTE` and `GRANT` actions to increase governance oversight.
-
-## [5.3.0] — Dynamic Blast Radius — 2026-03-28
-
-🚀 **MindForge v5.3.0 — Pillar II Implementation (APO v2)**
-
-This update introduces real-time risk assessment for agentic actions, preventing high-impact regressions through dynamic guardrails.
-
-### Pillar II: Dynamic Blast Radius (v5.3.0)
-
-- **Impact Scoring Engine**: Implemented `impact-analyzer.js` to calculate the "Blast Radius" (0-100) of every intent based on action severity and target sensitivity.
-- **Max-Impact Policy Enforcement**: Updated `policy-engine.js` to enforce `max_impact` thresholds, allowing granular control over destructive operations.
-- **Fail-Safe Security**: Integrated "Default Deny" posture if impact analysis fails, defaulting to score 100 (CRITICAL).
-
-## [5.2.0] — Semantic Vector Consensus — 2026-03-28
-
-🚀 **MindForge v5.2.0 — Pillar I Implementation (FIM v2)**
-
-This update transcends simple timestamp-based synchronization, moving to a vector-space consensus model for organizational knowledge.
-
-### Pillar I: Semantic Vector Consensus (v5.2.0)
-
-- **Hybrid Semantic Synthesis**: Refactored `federated-sync.js` to use `EmbeddingEngine` cosine similarity for knowledge merging.
-- **4-Branch Resolution Protocol**:
-    - **LWW (> 0.9)**: Auto-resolve near-identical entries.
-    - **Autonomous ADS Merge (0.75 - 0.9)**: Proactive synthesis of overlapping knowledge.
-    - **Nexus Handover (0.6 - 0.75)**: Human-agent conflict resolution.
-    - **Topic Isolation (< 0.6)**: Prevention of ID collisions for disparate content.
-- **Resilient Sync Architecture**: Hardened synchronization loop with robust error handling and TF-IDF fallback.
-
-## [5.1.0] — The Enterprise Expansion — 2026-03-28
-
-🚀 **MindForge v5.1.0 — Evolution of the Agentic Protocol Mesh**
-
-This major update integrates 14 advanced agentic protocols and high-performance session hooks originally developed in the Superpowers framework, fully sanitized and hardened for the MindForge ecosystem.
-
-### 🧩 Advanced Agentic Protocols (v5.1.0)
-
-- **Extended Tactical Rigor**: Overlapping skills have been evolved into `_extended` variants (e.g., `mindforge-debug_extended`, `mindforge-tdd_extended`) with higher verification standards.
-- **Neural Orchestration**: New `mindforge-neural-orchestrator` layer for intelligent protocol activation during session startup.
-- **Autonomous Multi-Tasking**: Integrated `parallel-mesh` and `swarm-execution` for mass-scale phase execution.
-- **Creative Intelligence**: New `mindforge:brainstorming` workflow for intent and architectural exploration.
-- **Hardened QA Gates**: Evolution of `verify-work` and `ship` protocols with automated quality assurance and branch sanitization.
-- **Isolated Workspaces**: Support for clean Git worktrees via `workspace-isolated`.
-
-### 🛠️ Core Infrastructure Updates (v5.1.0)
-
-- **Native Session Hooks**: Replaced legacy bash hooks with high-performance Node.js session-initialization drivers.
-- **Workflow Synchronization**: Core command suite (/plan, /execute, /debug, /ship etc.) updated to mandate Step 0 Protocol Activation.
-- **Full Sanitization**: 100% elimination of external branding across all documentation and system identifiers.
-
-### Added (v5.1.0)
-
-- `mindforge-brainstorming` (New)
-- `mindforge-parallel-mesh_extended` (Extended)
-- `mindforge-execute-phase_extended` (Extended)
-- `mindforge-ship_extended` (Extended)
-- `mindforge-review-inbound` (New)
-- `mindforge-review-request` (New)
-- `mindforge-swarm-execution` (New)
-- `mindforge-debug_extended` (Extended)
-- `mindforge-tdd_extended` (Extended)
-- `mindforge-workspace-isolated` (New)
-- `mindforge-neural-orchestrator` (New)
-- `mindforge-verify-work_extended` (Extended)
-- `mindforge-plan-phase_extended` (Extended)
-- `mindforge-skill-creation` (New)
-
----
-
-## [5.0.0-alpha.2] — Predictive Reliability & Supply Chain Trust — 2026-03-28
-
-🚀 **MindForge v5.0.0 — Pillars III & IV Implementation**
-
-This update adds the next two critical architectural pillars to the v5 Enterprise roadmap, focusing on long-term session reliability and a verifiable asset supply chain.
-
-### Pillar III: Predictive Agentic Reliability (PAR)
-
-- **Advanced Loop Detection**: Implemented **S03 (Semantic Mirroring)** and **S04 (Infinite Decomposition)** patterns in the `StuckMonitor`.
-- **Context Density Refactorer**: New proactive utility that triggers context summarization and handoffs when reasoning-to-action density falls below 30%.
-- **C2C Arbitrage**: Integrated **Confidence-to-Cost** threshold gating into the `AutoRunner` to prevent low-value autonomous drift.
-- **Self-Healing Integration**: Automated triggering of "hindsight injection" and state repair when PAR patterns are detected.
-
-### Pillar IV: Supply Chain Trust (ZTS)
-
-- **Agentic SBOM**: Implemented the `NexusTracer` manifest generator, producing real-time `MANIFEST.sbom.json` files for every reasoning chain.
-- **7-Dimension Certification (7D)**: Re-architected the `skill-validator` with a weighted scoring system covering Schema, Triggers, Security, Clarity, and Examples.
-- **Enterprise-Grade Enforcement**: Strict `--enterprise` mode requirement for a minimum **7.0/10.0** certification score.
-- **Model/Skill Telemetry**: Integrated real-time tracking of asset provenance and usage history in the audit stream.
-
-## [5.0.0-alpha.1] — Federated Intelligence & Policy Orchestration — 2026-03-28
-
-🚀 **MindForge v5.0.0 — The Sovereign Enterprise Release**
-
-This landmark release transforms MindForge into a distributed, governable, and cryptographically verified agentic operating system. V5 introduces the first two pillars of the Enterprise Architecture: Federated Intelligence Mesh (FIM) and Agentic Policy Orchestrator (APO).
-
-### Pillar I: Federated Intelligence Mesh (FIM)
-
-- **Enterprise Intelligence Service (EIS)**: Implemented `eis-client.js` for secure, authenticated communication with the organizational knowledge hub.
-- **Delta-Sync Protocol**: Implemented `federated-sync.js` with timestamp-based delta pulls, significantly reducing synchronization latency.
-- **ZTAI-Signed Provenance**: Every knowledge item in the mesh is cryptographically tied to the DID of the originating agent.
-- **Hybrid Knowledge Graph**: Extended `knowledge-graph.js` to seamlessly resolve local nodes and remote federated nodes.
-
-### Pillar II: Agentic Policy Orchestrator (APO)
-
-- **Policy-as-Code (PaC)**: Implemented `policy-engine.js` for intent-based evaluation of agent actions against organizational security rules.
-- **RBAC Manager**: Implemented `rbac-manager.js` for mapping DIDs to project roles and binding permissions to ZTAI Trust Tiers.
-- **Policy Interceptor**: Deep integration into `auto-runner.js`, enforcing a pre-flight governance gate before every autonomous wave.
-- **Default Enterprise Policies**: Shipped with initial security guardrails for engine and infrastructure protection.
-- **Dynamic Blast Radius (v5.3.0)**: Integrated `ImpactAnalyzer` to calculate and enforce risk-based thresholds for autonomous actions.
-
-### Hardening (Enterprise Mode v5.0.0-alpha.1)
-
-- **ZTAI Interlock**: All mesh and policy operations now utilize the hardware-enclave (simulated) signing engine for Tier 3 principals.
-- **Dynamic Intent Extraction**: Autonomous intents are now derived in real-time from active session identities.
-- **Conflict Resolution (v5.2.0)**: Uses **Hybrid Semantic Synthesis** (Pillar I v5.2.0). 
-  - **Similarity > 0.9**: Near-identical; auto-resolve via **Last-Write-Wins (LWW)**.
-  - **0.75 - 0.9**: Semantic Overlap; triggers **Autonomous ADS Merging**.
-  - **0.6 - 0.75**: Conflict; triggers **Nexus Handover (DHH)** for human steering.
-  - **< 0.6**: Topic Collision; isolates entries into unique IDs to prevent data loss.
-
----
-
-## [4.3.0] — Enterprise Mesh & Proactive Equilibrium — 2026-03-28
-
-🚀 **MindForge v4.3.0 — The Maturity Release**
-
-This major update completes the **MindForge V4 Roadmap**, introducing autonomous financial governance and self-healing reliability engines. It also includes a comprehensive structural reorganization of the framework for enterprise scalability.
-
-### Key Pillars (v4.3.0)
-
-- **Autonomous FinOps Hub (Pillar V)**: Implemented `ModelBroker` for dynamic **C2C (Confidence-to-Cost)** routing and `FinOpsHub` for real-time ROI tracking.
-- **Proactive Equilibrium (Pillar VI)**: Implemented `WaveFeedbackLoop` for divergence detection and `TemporalHindsight` for autonomous state recovery (RCA/Repair).
-- **Structural Reorganization**: Migrated all core engine implementation files from `.mindforge/` logic directories to a standardized `bin/` domain architecture.
-- **Global Memory Sync**: Enhanced `SemanticHub` to support organizational-level knowledge sharing via `~/.mindforge/`.
-- **Hardened ZTAI Enterprise Mode**: HSM-simulated signing for Tier 3 principals and automated audit integrity manifests.
-
-🚀 **MindForge v4.2.0 — The Security Revolution**
-
-This update introduces **Zero-Trust Agentic Identity (ZTAI)**, a core pillar of the MindForge enterprise security architecture. It establishes a DID-based identity system with cryptographic signing for all agentic actions, ensuring non-repudiable audit trails and tiered governance.
-
-### Key Pillars (v4.2.0)
-
-- **ZTAI Identity Engine**: Core manager for DID generation (`did:mindforge:<uuid>`) and Ed25519 key pair lifecycle.
-- **Cryptographic Audit Signing**: All reasoning traces and spans are signed by the originating agent, ensuring log integrity.
-- **Trust-Tiered Governance**: 4-tier authorization model (0-3) for granular permission control across the agentic mesh.
-- **Identity-Lock Engine**: Automated security gate in the `WaveExecutor` that blocks unauthorized Tier 3 actions.
-- **Trust Verifier**: Real-time cryptographic validation of the `AUDIT.jsonl` log for cross-session integrity.
-- **Enterprise-Grade Hardening**: session-based key rotation and hardware security module (HSM) hook placeholders.
-
----
-
-## [4.1.0] — MindForge Nexus (Observability & Tracing) — 2026-03-27
-
-🚀 **MindForge v4.1.0 — The Observability Revolution**
-
-This update introduces **MindForge Nexus**, a high-fidelity observability and tracing system designed for the agentic mesh. It enables deep visibility into agentic "thought chains" and parallel swarm execution through OpenTelemetry-compatible tracing.
-
-### Key Pillars (v4.1.0)
-
-- **Nexus Tracer Engine**: Core engine for managing hierarchical spans and trace context propagation across the mesh.
-- **Agentic Reasoning Tracing (ART)**: New audit event type `reasoning_trace` for capturing granular agentic thought cycles.
-- **Hierarchical Span Orchestration**: Universal ART fields (`trace_id`, `span_id`, `parent_span_id`) integrated into all audit events.
-- **Mesh-Integrated Tracing**: Direct injection of tracing protocols into the `Wave Executor` and `Swarm Controller`.
-- **Nexus Dashboard Spec**: Architectural blueprint for real-time trace visualization and reasoning heatmaps.
-- **ZTAI Observability**: Integrated trust-tier visibility for secure execution monitoring.
-
----
-
-## [4.0.0] — Dynamic Multi-Agent Swarms (Agentic Mesh) — 2026-03-27
-
-🚀 **MindForge v4.0.0 — The Mesh Revolution**
-
-This landmark release introduces **Dynamic Multi-Agent Swarms (Agentic Mesh)**, shifting MindForge from sequential subagent execution to a parallel, task-aware, and verifiable cluster-based orchestration system. V4 enables mass-scale autonomous engineering through ephemeral specialist clusters and zero-trust resonance.
-
-### Key Pillars (v4.0.0)
-
-- **Dynamic Swarm Orchestration**: The new `SwarmController` analyzes task complexity and spawns ephemeral clusters of specialist personas in parallel.
-- **Micro-Persona Factory**: On-the-fly persona specialization using task-specific knowledge injection via `Context7`.
-- **Wave Executor (Mesh Mode)**: Refactored parallel execution engine with shared `SWARM-STATE` synchronization and leader-led synthesis.
-- **ZTAI (Zero-Trust Agentic Identity)**: DID-based cryptographic signing for swarm actions and multi-tier trust enforcement.
-- **Enterprise specialist Library**: 12+ pre-defined swarm templates including `AIEngineeringSwarm`, `IdentityTrustSwarm`, `DataMeshSwarm`, and `IncidentResponseSwarm`.
-- **HITL Governance Gates**: Human-in-the-loop decision gates for high-risk (Tier 3) architectural and compliance-heavy operations.
-
----
-
-## [3.0.0] — Reactive Autonomous Intelligence — 2026-03-27
-
-🚀 **MindForge v3.0.0 — The Architecture Revolution**
-
-This major release transforms MindForge from a "disciplined workflow engine" into a **Reactive Autonomous Intelligence**. By sharding context via semantic relevance and debating architectural decisions through adversarial model loops, V3 achieves sub-second reasoning and zero logic-drift in complex engineering sessions.
-
-### Key Pillars (v3.0.0-rc1)
-
-- **Context Sharding (SRD)**: 40% reduction in token waste via relevance-dense Hot/Warm/Cold context tiers.
-- **Adversarial Decision Synthesis (ADS)**: Zero-drift architectural logic through Red-Team/Blue-Team debate and SOUL-scoring.
-- **Temporal Vision**: Full-fidelity history navigation, hindsight injection, and automated state repair.
-- **RAG 2.0 (Auto-Shadowing)**: Background pattern retrieval from the local knowledge graph without manual prompts.
-
----
-
-## [2.6.0] — Temporal Vision (Time-Travel Debugging) — 2026-03-27
-
-🚀 **MindForge v2.6.0 — The Observability Revolution**
-
-This update introduces "Temporal Vision," a high-fidelity time-travel debugging system that allows developers to scrub through execution history, explore past states, and perform "Hindsight Injection" to repair architectural drift automatically.
-
-### Added (v2.6.0)
-
-- **Temporal Vision Engine**:
-  - **Temporal Hub**: Synchronous state snapshotting of the `.planning/` directory at every significant audit point.
-  - **State Snapshots**: High-fidelity records of `.md`, `.json`, and `.log` files stored in `.planning/history/[audit_id]/`.
-  - **Hindsight Injection**: Automated state repair protocol that rolls back history to a specific point ($T_n$), injects a correction, and re-triggers autonomous execution.
-- **Observability Interface (Dashboard Integration)**:
-  - **Temporal API**: REST endpoints for history navigation, snapshot exploration, and remote hindsight injection.
-  - **Auto-Runner Integration**: Unified event ID generation and automated state capture embedded in the core loop.
-- **CLI Extensions**:
-  - `/mindforge:temporal status`: View snapshot density and history stats.
-  - `/mindforge:temporal inject`: Manual hindsight injection from the terminal.
-  - `/mindforge:temporal cleanup`: Automated purge of stale history snapshots.
-- **Governance**:
-  - **Temporal Protocol (`temporal-protocol.md`)**: Definition of safety rules, recovery boundaries, and state integrity checks.
-
----
-
-## [2.5.0] — Local-First Knowledge Graph (RAG 2.0) — 2026-03-27
-
-🚀 **MindForge v2.5.0 — The Knowledge Graph Revolution**
-
-This update introduces a sophisticated Local-First Knowledge Graph (RAG 2.0), moving beyond flat vector search to a typed-edge, traversable knowledge architecture with proactive context shadowing.
-
-### Added (v2.5.0)
-
-- **Local-First Knowledge Graph (RAG 2.0)**:
-  - **Embedding Engine**: Custom local TF-IDF vectorizer with bigram detection and camelCase splitting. Zero external dependencies.
-  - **Graph Engine**: Typed-edge graph (6 types) with BFS traversal, cycle detection, and SHA-256 edge integrity.
-  - **Auto-Shadow Engine**: Proactive "ghost-pattern" context injection for subagents based on graph proximity and semantic similarity.
-  - **Typed Edges**: Support for `RELATED_TO`, `CAUSED_BY`, `SUPERSEDES`, `DEPENDS_ON`, `INFORMS`, and `CONTRADICTS` relationships.
-  - **Graph-Aware Capture**: Automatic edge inference during memory capture from bug patterns and review findings.
-  - **Maintenance Tools**: Integrity verification, edge weight decay (time-based), and orphan node pruning.
-- **SDK Extensions**:
-  - Full TypeScript support for graph traversal, edge management, and hybrid similarity queries.
-
-### Fixed (v2.5.0)
-
-- **macOS App Sandbox Compatibility**: Implemented `setTestMode()` in memory engines to bypass nesting-related EPERM errors when running from sandboxed editors.
-
----
-
-## [2.4.0] — The Core Pillars Optimization — 2026-03-27
-
-🚀 **MindForge v2.4.0 — The Core Pillars Optimization**
-
-This major architectural update introduces a dynamic, SRD-based memory sharding system and the Adversarial Decision Synthesis (ADS) protocol, replacing legacy monolithic handoffs with a hardened, high-fidelity Tri-Tier Memory model.
-
-### Added (v2.4.0)
-
-- **Adversarial Decision Synthesis (ADS)**:
-  - **ADS Engine**: Integrated a 3-model synthesis loop (Architect/Auditor/Synthesizer) for peer-reviewed architectural planning.
-  - **SOUL Scoring Engine**: New scoring metric `(Impact * Leverage * Reversibility) / (Effort * Risk * Cost)` for objective decision ranking.
-  - **Red-Team Jailbreak**: Enforced architectural auditing that forces the identification of at least 3 critical system flaws.
-  - **Automated ADR Workflow**: synthesis results are now persisted as SOUL-scored Decision Records in `.planning/decisions/`.
-- **Semantic Context Sharding**:
-  - **Tri-Tier Memory Controller**: Integrated `shard-controller.md` to manage memory transitions across Hot (`HANDOFF.json`), Warm (`.planning/memories/`), and Cold (`.mindforge/memory/`) tiers.
-  - **SRD Scoring Engine**: New weighted Semantic Relevance Density (SRD) scoring for contextual items based on Decisiveness, Frequency, and Impact.
-  - **Enterprise-Grade Hardening**: Implemented SHA-256 integrity verification and automated semantic tagging for all memory shards.
-  - **Proactive Warm Retrieval**: Updated the context injector to automatically pull relevant shards based on task sub-context.
-- **Shard Helper Utility**: New `bin/shard-helper.js` for standalone memory analysis and integrity auditing.
-
-### Changed (v2.4.0)
-
-- **Planning Protocol**: `/mindforge:plan-phase` now officially supports the `--ads` flag for high-fidelity synthesis cycles.
-- **Compaction Protocol V3**: Upgraded the standard compaction flow to include `Step 4.5: Semantic Sharding`, preventing long-term context window pollution and reducing context overhead by ~35%.
-
----
-
-## [2.3.5] — Intelligent Asset Sync & Merge — 2026-03-26
-
-🚀 **MindForge v2.3.5 — Intelligent Asset Sync & Merge**
-
-This release transforms the installation engine from a static "skip if exists" model to an intelligent "Merge & Sync" system, ensuring all framework assets, documentation, and advanced modules are correctly deployed even into existing project environments.
-
-### Added (v2.3.5)
-
-- **Intelligent Merging**: Refactored `copyDir` to support `noOverwrite` mode, allowing the installer to add missing subfolders and templates without disrupting user-customized files.
-- **Advanced Module Sync**: Added explicit path mapping for `memory`, `plugins`, `intelligence`, and `dashboard` modules into the standard installation payload.
-- **Expanded Asset Types**: Upgraded runtime definitions to include `memorySubdir` and `pluginsSubdir` for all supported IDE environments.
-
-### Fixed (v2.3.5)
-
-- **Sync Gating**: Removed restrictive directory existence checks that caused the installer to skip entire asset categories if the parent folder already existed.
-- **Documentation Parity**: Hardened the documentation sync logic to ensure `references/` and `templates/` always reach the `.agents/docs/` target.
-
----
-
-## [2.3.4] — Architectural Sync & Zero-GSD Purge — 2026-03-25
-
-### Added (v2.3.4)
-
-- **Comprehensive Planning Sync**: Expanded `.planning` deployment to include `ROADMAP.md`, `ARCHITECTURE.md`, `REQUIREMENTS.md`, and `RELEASE-CHECKLIST.md` by default.
-- **Zero-GSD Standard**: Completed the removal of all "GSD" instances from the installer logic and internal system identifiers.
-
-### Fixed (v2.3.4)
-
-- **Documentation Payload**: Fixed a discrepancy in the path-mapping for documentation subdirectories (`references/`, `templates/`) ensuring they correctly sync into `.agents/docs/`.
-- **Directory Persistence**: Ensured that empty planning subdirectories correctly persist with `.gitkeep` during fresh installations.
-
----
-
-## [2.3.3] — Payload Cleanup & Enterprise Pruning — 2026-03-25
-
-🚀 **MindForge v2.3.3 — Payload Cleanup & Enterprise Pruning**
-
-This release strictly prunes the installation payload to remove legacy GSD artifacts, project-specific state, and development-only folders, ensuring a clean "out-of-the-box" enterprise experience.
-
-### Fixed (v2.3.3)
-
-- **Installation Payload**: Removed legacy migration folders (`01-migrate-legacy-to-mindforge`, `day1`, `day2`, `day3`).
-- **State Leakage**: Excluded project-specific state files (`AUDIT.jsonl`, `HANDOFF.json`, `jira-sync.json`, `slack-threads.json`) from distribution.
-- **Enterprise Pruning**: Updated `installer-core.js` to automatically filter out development-only framework folders (e.g., `distribution`, `monorepo`, `production`).
-
----
-
-## [2.3.2] — Asset Sync Repair — 2026-03-25
-
-🚀 **MindForge v2.3.2 — Asset Sync Repair**
-
-This release fixes documentation asset synchronization and repairs logic corruption in the installer core.
-
-### Added (v2.3.2)
-
-- **Docs & Templates Sync**: Integrated `docs/references` and `docs/templates` into the standard installation payload.
-- **Recursive Counting**: Implemented deep directory file counting for accurate manifest reporting.
-
-### Fixed (v2.3.2)
-
-- **Installer Logic**: Repaired syntax corruption and removed duplicated code blocks in `installer-core.js`.
-
----
-
-## [2.1.2]*Status: V5.2.0 Semantic Consensus Implemented & Verified (2026-03-28)*
-
-🚀 **MindForge v2.1.2 — Enterprise Branding & CI Fix**
-
-This release transforms the MindForge installation experience with a high-impact "Enterprise-Grade" CLI interface and resolves critical CI/CD integration issues.
-
-### Added (v2.1.2)
-
-- **Enterprise CLI Branding**: New ultra-blocky ASCII "MINDFORGE" branding and high-contrast "Digital Architect" design system.
-- **Payload Manifest Summary**: Architectural summary screen displaying dynamic counts of deployed Personas, Skills, and Integrations.
-- **"TRY IT NOW" Utility**: Boxed terminal widget providing immediate post-install command guidance.
-- **Enhanced Status Reporting**: Unified architectural status grid for environment detection.
-
-### Fixed (v2.1.2)
-
-- **CI/CD Permissions**: Resolved `RequestError [HttpError]: GitHub Actions is not permitted to create or approve pull requests` by adding `contents: write` to `auto-pr.yml`.
-
----
-
-## [2.1.1] — Core Migration Finalization — 2026-03-25
-
-🧩 **MindForge v2.1.1 — Core Migration Finalization**
-
-This release completes the structural rebranding and migration of the MindForge framework into MindForge. It finalizes the path alignment for all 120+ assets, hardens the core configuration management, and fully integrates the expanded persona ecosystem.
-
-### Added (v2.1.1)
-
-- **Finalized Path Mapping**: All framework assets (skills, workflows, bin) now reside in a unified, flat `.agent/` structure for cross-IDE compatibility.
-- **32-Persona Ecosystem**: Full integration of 32 specialized engineering personas with defined tool permissions and capability matrices.
-- **Unified 4-Pillar Workflow**: Hardened `plan`, `execute`, `verify`, and `ship` logic with project-level `.agent/` integrity.
-- **Enterprise Manifest**: New `file-manifest.json` serving as the single source of truth for multi-project codebase mapping.
-- **Consolidated Configuration**: Unified framework settings and project-level governance in `.mindforge/` and `.agent/`.
-- **Rebranded Lifecycle Hooks**: 5 new high-performance hooks for context monitoring, prompt guarding, and real-time status updates.
-
-### Changed (v2.1.1)
-
-- **Zero-Watermark Integrity**: 100% elimination of residual "MindForge" and "get-shit-done" identifiers from codebase, documentation, and metadata.
-- **Documentation Overhaul**: Modernized the entire documentation suite, including Architecture, Personas, Skills Authoring, and Command References.
-- **Hardened Settings**: `settings.json` now features dynamic hook pathing for increased portability across system environments.
-
----
-
-## [2.1.0] — The Specialized Expansion — 2026-03-24
-
-🧩 **MindForge v2.1.0 — The Specialized Expansion**
-
-This release significantly expands the MindForge persona ecosystem, integrating 18+ high-performance specialized agents from the MindForge framework. This expansion doubles the framework's baseline capabilities in research, architecture, execution, and quality assurance.
-
-### Added (v2.1.0)
-
-- **18+ New Specialized Personas**: Integration of advanced agents including `nyquist-auditor`, `user-profiler`, `advisor-researcher`, `ui-auditor`, and more.
-- **Extended Mapping & Strategy**: Enhanced `roadmapper-extend` and `codebase-mapper-extend` for deep project lifecycle management.
-- **Comprehensive Permissions Audit**: Full tool-access transparency for all 32 personas in `docs/PERSONAS.md`.
-- **Architectural State Update**: `Master-Context.md` now reflects the 32-persona structural foundation.
-
-### Changed (v2.1.0)
-
-- **Documentation Overhaul**: Complete update of `PERSONAS.md` and `Master-Context.md` to reflect the expanded ecosystem.
-- **Branding Alignment**: All integrated MindForge personas rebranded and fully sanitized for MindForge native use.
-
----
-
-## [2.0.0] — The Autonomous Enterprise — 2026-03-24
-
-🚀 **MindForge v2.0.0 — The Autonomous Enterprise — Major Release**
-
-This major release transforms MindForge from a Claude-centric framework into a universal AI agent operating system. With support for 6 major runtimes and a hardened autonomous execution engine, v2.0.0 is built for enterprise-scale AI orchestration.
-
-### Added (v2.0.0)
-
-- **6-Runtime Multi-Platform Support**: Official support for Claude Code, Antigravity, Cursor, OpenCode, Gemini CLI, and GitHub Copilot.
-- **Runtime Adapters**: Automatic content transformation for Gemini (model/filename mapping) and preambles for non-slash runtimes (Cursor/Copilot).
-- **`/mindforge:new-runtime`**: Rapidly scaffold custom AI agent runtimes with correct directory structures and command visibility.
-- **Unified Migration Engine (v2.0.0)**: Additive schema upgrades for `AUDIT.jsonl` (runtime/agent_id) and `token-usage.jsonl` (model_group).
-- **Hardened Self-Building Skills**: Automated skill capture from documentation and phase outputs.
-- **7-Dimension Quality Scorer**: Enhanced static analysis for skill authoring.
-- **MindForge Integration**: 10 new advanced commands for design-first planning, zero-friction capture, and smart routing.
-- **`/mindforge:do`**: Smart natural language dispatcher that routes intent to the correct command.
-- **`/mindforge:ui-phase` & `/mindforge:ui-review`**: Tier-1 visual design contract and visual audit engine.
-- **`/mindforge:note`**: Instant idea capture with todo promotion.
-- **`/mindforge:validate-phase`**: Requirement coverage and test gap identification.
-- **`/mindforge:session-report`**: Professional stakeholder summary with token profiling.
-- **Backlog & Seed Management**: `/mindforge:add-backlog`, `/mindforge:review-backlog`, and `/mindforge:plant-seed`.
-- **Parallel Workstreams**: `/mindforge:workstreams` for concurrent milestone execution.
-- **65-Point Production Checklist**: Exhaustive verification suite for enterprise readiness.
-
-### Changed (v2.0.0)
-
-- **Directory Standard**: Antigravity local workflows moved to `.agents/workflows/` with mandatory YAML frontmatter.
-- **Migration Strategy**: Transitioned to an idempotent, backup-first migration mod- **ADR-039**: Multi-Runtime Platform Support.
-- **ADR-040**: Additive Schema Migration Strategy.
-- **ADR-041**: Stable Runtime Interface Contract.
-
----
-
-## [2.0.0-alpha.12] — Self-Building Skills Platform — 2026-03-24
-
-### Added (v2.0.0-alpha.12)
-
-- **Self-Building Skills Platform**: Intelligent engine for automatically capturing skills from documentation, sessions, and npm packages.
-- **`/mindforge:learn`**: High-level orchestration command for 7-step skill generation pipeline.
-- **`/mindforge:marketplace`**: Seamless integration with npm registry for community skills discovery and installation.
-- **7-Dimension Quality Scorer**: Static analysis tool ensuring high-quality, injection-safe skill authoring.
-- **Auto-Capture Hook**: Automatic pattern detection from phase outputs (SUMMARY, ADR, HANDOFF).
-- **CLI Commands**: `learn-skill` and `marketplace` added to `bin/mindforge-cli.js`.
-- **ADR-036, ADR-037, ADR-038**: New architectural decisions for skill sourcing, capture thresholds, and quality gates.
-
----
-
-## [2.0.0-alpha.11] — Persistent Knowledge Graph — 2026-03-24
-
-### Added (v2.0.0-alpha.11)
-
-- **Persistent Knowledge Graph**: Core engine for long-term memory across projects.
-- **`remember` CLI**: New command for manual knowledge management and semantic search.
-- **Automated Capture**: Hooks in `AutoRunner` and `CrossReviewEngine` to ingest insights automatically from phase completions and review reports.
-- **Context Injection**: `installer-core.js` now auto-injects relevant historical context into `CLAUDE.md` during installation.
-- **`test-memory` CLI**: New command for memory suite verification.
-
----
-
-## [2.0.0-alpha.10] — Skill Installation & Validation CLI — 2026-03-24
-
-### Added (v2.0.0-alpha.10)
-
-- New CLI commands: `validate-skill`, `install-skill`, `register-skill`, `audit-skill`.
-- Robust skill validation logic (Level 1 & 2) in `bin/skill-validator.js`.
-- Automated manifest management and audit logging in `bin/skill-registry.js`.
-- Integrated CLI commands into the `install-skill.md` agentic workflow for higher reliability.
-
----
-
-## [2.0.0-alpha.9] — Verified Antigravity Compatibility — 2026-03-23
-
-### Fixed (v2.0.0-alpha.9)
-
-- Fixed critical installer regression where `files` variable was undefined during command installation.
-- Fixed `resolveBaseDir` hardcoding that prevented `.agents/ relocation in non-standard environments.
-- Verified end-to-end Antigravity fix for command visibility and YAML frontmatter.
-
----
-
-## [2.0.0-alpha.8] — Antigravity Visibility Fix — 2026-03-23
-
-### Fixed (v2.0.0-alpha.8)
-
-- **Antigravity Command Visibility**: Relocated workflows to `.agents/workflows/` (from `agents/`) and injected mandatory YAML frontmatter (`description`) to ensure registration in Antigravity.
-- **Zero-Config Sync**: Updated `init-project` command to reflect the new `.agents` directory standard.
-
----
-
-## [2.0.0-alpha.7] — Day 12: Real-time Observability Dashboard — 2026-03-22
-
-### Added (v2.0.0-alpha.7)
-
-- [NEW] Real-time Dashboard Server (Express + SSE Bridge).
-- [NEW] Premium 5-tab UI: Activity, Metrics, Approvals, Memory, Team.
-- [NEW] Live Data Streaming for Audit Logs, Quality, Costs, and Team Activity.
-- [NEW] Hardened Tier 3 Governance: Mandatory Plan ID typing for sensitive approvals.
-- [NEW] Metrics Aggregator: Multi-source JSONL processing for observability.
-- [NEW] CLI Command: `/mindforge:dashboard` for server lifecycle management.
-- [NEW] ADR-033: Real-time Observability Architecture.
-- [HARDENED] SSE Rotation Detection (Inode-based).
-- [HARDENED] Localhost-only Security Model (Strict CORS/Binding).
-
----
-
-## [2.0.0-alpha.4] — Day 11: Persistent Knowledge Graph (Long-Term Memory) — 2026-03-22
-
-### Added (v2.0.0-alpha.4)
-
-- [NEW] Persistent Knowledge Graph for long-term memory across sessions.
-- [NEW] TF-IDF search engine with confidence-based relevance ranking.
-- [NEW] Automated Knowledge Capture from ADRs, Debug Reports, and Retrospectives.
-- [NEW] Global Knowledge Sync for machine-wide insight sharing (`~/.mindforge/global-knowledge-base.jsonl`).
-- [NEW] Skill SDK integration for programmatic memory access.
-- [NEW] Manual memory management command (`/mindforge:remember`).
-- [NEW] Append-only audit-ready storage schema for all knowledge types.
-- [NEW] ADR-030, ADR-031, ADR-032.
-
----
-
-## [2.0.0-alpha.3] — Day 10: Multi-Model Intelligence Layer — 2026-03-22
-
-### Added (v2.0.0-alpha.3)
-
-- [NEW] Unified Multi-Model Intelligence Layer (Anthropic, OpenAI, Gemini).
-- [NEW] Dynamic Model Router with Persona and Tier-based logic.
-- [NEW] Real-time Cost Tracking and Daily Budget Enforcement.
-- [NEW] Adversarial Cross-Model Code Review Engine (`/mindforge:cross-review`).
-- [NEW] Deep Research Engine with Gemini 1.5 Pro 1M context support (`/mindforge:research`).
-- [NEW] Model usage and cost reporting (`/mindforge:costs`).
-- [NEW] Secure API handling with SSRF protection and header-based auth.
-- [NEW] ADR-027, ADR-028, ADR-029.
-
----
-
-## [2.0.0-alpha.2] — Day 9: Visual QA Engine (Browser Runtime Engine) — 2026-03-22
-
-### Added (v2.0.0-alpha.2)
-
-- [NEW] Persistent Browser Runtime (Playwright-powered Chromium daemon).
-- [NEW] Visual Verification Hook (`<verify-visual>`) in PLAN files.
-- [NEW] Systematic Visual QA Engine (`/mindforge:qa`).
-- [NEW] Persistent Session Management for Browser Control.
-- [NEW] ADR-024, ADR-025, ADR-026.
-
----
-
-## [2.0.0-alpha.1] — Day 8: Autonomous Execution Engine — 2026-03-22
-
-### Added (MindForge v2.0.0-alpha.1)
-
-**Autonomous execution engine:**
-
-- `/mindforge:auto` — walk-away autonomous phase/milestone execution
-- `/mindforge:steer` — mid-execution guidance injection from second terminal
-- Node repair operator: RETRY → DECOMPOSE → PRUNE → ESCALATE
-- Stuck detection engine: 5 patterns (S01-S05)
-- Headless CLI mode: `mindforge-cc headless --phase N`
-- `.planning/steering-queue.jsonl` — steering instruction queue
-- `.planning/auto-state.json` — real-time execution state
-
-**MINDFORGE.md v2 settings:**
-
-- AUTO_MODE_DEFAULT_TIMEOUT_MINUTES, AUTO_MODE_UAT
-- AUTO_NODE_REPAIR_BUDGET, AUTO_RETRY_ON_VERIFY_FAIL
-- AUTO_TASK_MAX_TOKENS, AUTO_TASK_TIMEOUT_MINUTES
-- AUTO_PUSH_ON_WAVE_COMPLETE, AUTO_NOTIFY_ON_ESCALATION
-
-### Hardened (v2.0.0-alpha.1)
-
-- Gate 3 (secret detection) now runs PRE-COMMIT in auto mode
-- Pre-flight dirty check excludes `.planning/` state files
-- DECOMPOSE:- **Deny**: The action is blocked, and the violation is logged to `AUDIT.jsonl`.
-- **Blast Radius Denial (v5.3.0)**: Action is blocked if the `Impact Score` exceeds the policy `max_impact` threshold.
-- **Escalate**: The action requires a higher-tier DID signature or explicit HITL (Human-in-the-Loop) approval.
-eserves module/package names
-- Steering injection guard validates all instructions
-- SIGTERM handler waits for task cleanup before saving state
-
-### Architecture Decisions (v2.0.0-alpha.1)
-
-- ADR-021: Autonomy Boundary
-- ADR-022: Node Repair Hierarchy
-- ADR-023: Gate 3 Timing
-
----
-
-## [1.0.5] — v1.0.5 Minimal Install Option — 2026-03-22
-
-### Added (v1.0.5)
-
-- `--minimal` flag to install only essential project scaffolding.
-
----
-
-## [1.0.4] — v1.0.4 Antigravity Install Fix — 2026-03-22
-
-### Fixed (v1.0.4)
-
-- Antigravity local install now correctly copies commands and CLAUDE.md into `agents/`.
-
----
-
-## [1.0.3] — v1.0.3 Antigravity Agents Folder — 2026-03-22
-
-### Changed (v1.0.3)
-
-- Local Antigravity installs now target `agents/` by default (legacy `.agent/` detected and supported).
-
----
-
-## [1.0.2] — v1.0.2 CLI Bin Fix — 2026-03-22
-
-### Fixed (v1.0.2)
-
-- npm bin entry now uses a path format accepted by npm publish.
-
----
-
-## [1.0.1] — v1.0.1 Installer and Packaging Fixes — 2026-03-22
-
-### Fixed (v1.0.1)
-
-- Interactive setup now uses installer-core directly (no recursive wizard call).
-- Package bin entry corrected to use `mindforge-cc` → `bin/install.js`.
-
-### Changed (v1.0.1)
-
-- Added publish whitelist to reduce package size and exclude build-only files.
-
----
-
-## [1.0.0] — v1.0.0 First Stable Public Release — 2026-03-22
-
-🎉 **MindForge v1.0.0 — Enterprise Agentic Framework — First Stable Release**
-
-Release published: v1.0.0 (GitHub Releases).
-
-Built over 7 focused sprints, MindForge transforms Claude Code and Antigravity
-from powerful-but-unstructured AI tools into production-grade engineering
-partners with full governance, observability, and enterprise integration.
-
-### What ships in v1.0.0
-
-- **36 commands** across 7 workflow categories
-- **10 core skill packs** with three-tier registry (Core/Org/Project)
-- **8 specialised agent personas** covering all engineering roles
-- **Wave-based parallel execution** with dependency graph and automatic compaction
-- **Enterprise integrations**: Jira, Confluence, Slack, GitHub, GitLab
-- **Three-tier governance**: Tier 1 (auto) / Tier 2 (peer review) / Tier 3 (compliance)
-- **Five non-bypassable compliance gates** (secret detection, CRITICAL findings, tests, CVEs, GDPR)
-- **Intelligence layer**: health engine, difficulty scorer, anti-pattern detector, team profiling
-- **Public skills registry**: npm-based `mindforge-skill-*` ecosystem
-- **CI/CD integration**: GitHub Actions, GitLab CI, Jenkins adapters
-- **@mindforge/sdk**: TypeScript SDK with client, event stream, and command builders
-- **Monorepo support**: npm/pnpm/Nx/Turborepo/Lerna workspace detection
-- **AI PR Review**: Claude API-powered code review with context loading
-- **Self-update mechanism**: version check, changelog diff, scope-preserving apply
-- **Version migration engine**: schema migration from v0.1.0 through v1.0.0
-- **Plugin system**: extensible via `mindforge-plugin-*` npm namespace
-- **Token usage optimiser**: profiling and efficiency strategies
-- **50-point production readiness checklist**: fully verified before this release
-
-**20 Architecture Decision Records** documenting every major design choice
-
-- **15 test suites** with 3× consecutive run requirement
-- **Complete reference documentation**: commands, security, ADR index, threat model
-
-### Stable interface contract (v1.0.0)
-
-See ADR-020. All 36 commands, HANDOFF.json schema, AUDIT event types,
-@mindforge/sdk exports, and plugin.json format are stable in 1.x.x.
-
-### Breaking changes from 0.6.0 (v1.0.0)
-
-- VERIFY_PASS_RATE_WARNING_THRESHOLD in MINDFORGE.md is now 0.0-1.0 (was 0-100)
-  Run `/mindforge:migrate` to auto-convert
-- AUDIT.jsonl session_id field is now required (auto-backfilled by migration)
-- HANDOFF.json plugin_api_version field required for plugin compatibility
-
-### Installation (v1.0.0)
-
-```bash
-npx mindforge-cc@latest
-# or
-npx mindforge-cc@1.0.0 --claude --global
-```
-
----
-
-## [0.6.0] — Day 6 Distribution Platform
-
-### Added (v0.6.0)
-
-- Public skills registry: `npx mindforge-skills install/publish/search` (npm-based)
-- Skill validator: 3-level validation schema (schema, content, quality)
-- MINDFORGE.md JSON Schema: validation with non-overridable field markers
-- MindForge CI mode: GitHub Actions / GitLab CI / Jenkins integration
-- GitHub Actions workflow: health, security, quality, AI review jobs
-- AI PR Review Engine: Claude API-powered code review with context loading
-- Monorepo/workspace support: npm/pnpm/Nx/Turborepo/Lerna detection
-- Cross-package planner: topological execution order for monorepo phases
-- @mindforge/sdk: TypeScript SDK with client, event stream, and command builders
-- SSE event stream: real-time progress events via Server-Sent Events
-- /mindforge:init-org — organisation-wide MindForge setup command
-- /mindforge:install-skill — install skill from public/private registry
-- /mindforge:publish-skill — publish skill to npm registry
-- /mindforge:pr-review — AI code review powered by Claude API
-- /mindforge:workspace — monorepo workspace management
-- /mindforge:benchmark — skill effectiveness benchmarking
-- 3 new ADRs: ADR-015 npm registry, ADR-016 CI timeout, ADR-017 localhost SDK
-
-### Hardened (v0.6.0)
-
-- Registry: TOCTOU-safe temp directory (chmod 700), tarball size verification
-- CI: timeout exits with code 0 (soft stop), clear Tier 3 block messages
-- SDK: localhost-only SSE binding, Linux inotify fallback
-- AI review: robust daily limit (parse error tolerant), file-based diff truncation
-- Monorepo: affected package detection uses manifest paths (not depth assumption)
-
----
-
-## [0.5.0] — Day 5 Intelligence Layer
-
-### Added (v0.5.0)
-
-- framework health engine with 7-category diagnostics and safe auto-repair guidance
-- smart context compaction with Level 1/2/3 protocols and structured handoff extraction
-- phase difficulty scorer with weighted composite and task-granularity feedback
-- anti-pattern detection engine with false-positive controls for C01/B03/D01
-- skill gap analyzer for pre-planning capability coverage checks
-- team profile system and per-developer profile templates
-- metrics schema and quality tracker (`session`, `phase`, `skill usage`, `compaction quality`)
-- project constitution file: `MINDFORGE.md`
-- setup wizard modules: environment detection, idempotent config generation, interactive flow
-- 4 new commands: `/mindforge:health`, `/mindforge:retrospective`,
-  `/mindforge:profile-team`, `/mindforge:metrics`
-- 3 ADRs: ADR-012, ADR-013, ADR-014
-- new Day 5 test suites: `tests/intelligence.test.js`, `tests/metrics.test.js`
-
-### Changed (v0.5.0)
-
-- CLAUDE runtime contract now includes Day 5 intelligence behavior and
-  non-overridable MINDFORGE enforcement
-- package entrypoint now points to setup wizard
-
-### Fixed (v0.5.0)
-
-- AUDIT corruption handling now quarantines invalid lines instead of deleting entries
-- wizard reliability improvements: stdin TTY fallback, end-of-flow credential guidance,
-  and idempotent config replacement behavior
-
----
-
-## [0.4.0] — Day 4 Enterprise Integrations and Governance
-
-### Added (v0.4.0)
-
-- enterprise integration specs for Jira, Confluence, Slack, GitHub, and GitLab
-- integration connection manager with credential hygiene and retry guidance
-- governance layer: change classification, approvals, and compliance gates
-- multi-developer handoff and session merger protocols
-- 6 new commands: `/mindforge:audit`, `/mindforge:milestone`,
-  `/mindforge:complete-milestone`, `/mindforge:approve`,
-  `/mindforge:sync-jira`, `/mindforge:sync-confluence`
-- Day 4 test suites: `tests/integrations.test.js`, `tests/governance.test.js`
-- enterprise setup and governance guide docs
-
-### Changed (v0.4.0)
-
-- README and CLAUDE entrypoint updated for enterprise governance workflows
-- audit schema expanded for integration and governance events
-
-### Fixed (v0.4.0)
-
-- Jira transition handling now documents dynamic transition lookup instead of
-  hardcoded IDs
-- governance tiering now covers code-content matches and audit-history escalation
-
----
-
-## [0.3.0] — Day 3 Skills Platform
-
-### Added (v0.3.0)
-
-- 5 new core skill packs: performance, accessibility, data-privacy,
-  incident-response, database-patterns
-- Skills distribution engine: registry, loader, versioning, conflict resolver
-- 5 new commands: /mindforge:skills, /mindforge:review, /mindforge:security-scan,
-  /mindforge:map-codebase, /mindforge:discuss-phase
-- Persona customisation override system (project and phase level)
-- Skills Manifest (MANIFEST.md) with tier-based registration
-- Skills Authoring Guide for creating org and project skills
-- Injection guard for Tier 2/3 skill validation
-
-### Changed (v0.3.0)
-
-- execute-phase now uses multi-tier skills loading
-- plan-phase now reads CONTEXT.md from discuss-phase if available
-- CLAUDE.md updated with skills platform and new command awareness
-
-### Fixed (v0.3.0)
-
-- cursor pagination correctness in database-patterns skill (compound cursor)
-
----
-
-## [0.2.0] — Day 2 Wave Engine
-
-### Added (v0.2.0)
-
-- Wave-based parallel execution engine
-- Dependency parser and wave grouper
-- Context injector with minimum-context principle
-- Compaction protocol (automated at 70% context)
-- AUDIT.jsonl append-only pipeline with full schema
-- 4 new commands: /mindforge:next, /mindforge:quick, /mindforge:status, /mindforge:debug
-
----
+## Older releases
 
-## [0.1.0] — Day 1 Foundation
-
-### Added (v0.1.0)
-
-- Core directory scaffold
-- CLAUDE.md agent entry point
-- 8 agent persona definitions
-- 5 initial core skill packs
-- 6 slash commands: help, init-project, plan-phase, execute-phase, verify-phase, ship
-- npm installer (npx mindforge-cc)
-- State management: STATE.md, HANDOFF.json
-- Org context templates: ORG.md, CONVENTIONS.md, SECURITY.md, TOOLS.md
-persona definitions
-- 5 initial core skill packs
-- 6 slash commands: help, init-project, plan-phase, execute-phase, verify-phase, ship
-- npm installer (npx mindforge-cc)
-- State management: STATE.md, HANDOFF.json
-- Org context templates: ORG.md, CONVENTIONS.md, SECURITY.md, TOOLS.md
+See [changelogs/](./changelogs/) for the complete archive (all 98 versions back to v0.1.0), or [changelogs/README.md](./changelogs/README.md) for an indexed table of contents.