@sparkleideas/claude-flow-patch 3.1.0-alpha.44.patch.4 → 3.1.0-alpha.44.patch.6

package/AGENTS.md

@@ -1,6 +1,6 @@
 # @sparkleideas/claude-flow-patch
 
-> Runtime patches for `@claude-flow/cli` v3.1.0-alpha.40, `ruvector`, and `ruv-swarm` v1.0.20
+> Runtime patches for `@claude-flow/cli` v3.1.0-alpha.41, `ruvector`, and `ruv-swarm` v1.0.20
 
 **Claude Code instructions are in [CLAUDE.md](CLAUDE.md). Project-specific instructions (defect index, workflows, policies) are in [README.md](README.md).**
 
@@ -95,7 +95,7 @@ Use `$skill-name` syntax to invoke:
 
 [optional body]
 
-Co-Authored-By: claude-flow <ruv@ruv.net>
+Sparkling Ideas <henrik@sparklingideas.co.uk>
 ```
 
 Types: `feat`, `fix`, `docs`, `style`, `refactor`, `perf`, `test`, `chore`

package/CLAUDE.md

@@ -24,6 +24,7 @@ Runtime patches for `@claude-flow/cli` **v3.1.0-alpha.41**, `ruvector`, and `ruv
 - NEVER reuse a defect ID that was previously assigned to a different GitHub issue
 - ONE defect directory and ONE fix.py per GitHub issue -- do not combine multiple GitHub issues into one defect or split one GitHub issue across multiple defects
 - ALWAYS verify with `bash check-patches.sh` after applying
+- ALWAYS run `npm run preflight` before staging — the pre-commit hook (`hooks/pre-commit`) enforces this via `--check` mode
 - ALWAYS update ALL listing files when adding/removing a defect (see checklist)
 - Execution order is determined by the numeric prefix on each defect directory name. Dependencies between defects are expressed by assigning lower numbers to prerequisites.
 
@@ -38,7 +39,7 @@ lib/
   discover.mjs          # Dynamic discovery: scans patch/*/ → structured JSON
   categories.json       # Prefix-to-label mapping (one line per category)
 scripts/
-  update-docs.mjs       # Regenerates doc tables from discovery (npm run update-docs)
+  preflight.mjs         # Pre-commit sync: doc tables, versions, config (npm run preflight)
   upstream-log.mjs      # Show recent upstream releases (npm run upstream-log [count])
 patch/
   {NNN}-{PREFIX}-{NNN}-{slug}/    # NNN = 3-digit execution order
@@ -135,7 +136,7 @@ Save the returned GitHub issue number for the defect README.md.
 | DM | Daemon & Workers | 6 |
 | EM | Embeddings & HNSW | 2 |
 | GV | Ghost Vectors | 1 |
-| HK | Hooks | 3 |
+| HK | Hooks | 4 |
 | HW | Headless Worker | 4 |
 | IN | Intelligence | 1 |
 | MM | Memory Management | 1 |
@@ -145,7 +146,7 @@ Save the returned GitHub issue number for the defect README.md.
 | SG | Settings Generator | 2 |
 | UI | Display & Cosmetic | 2 |
 
-## All 31 Defects
+## All 32 Defects
 
 | ID | GitHub Issue | Severity |
 |----|-------------|----------|
@@ -162,9 +163,10 @@ Save the returned GitHub issue number for the defect README.md.
 | HK-001 | [#1155 post-edit hook records file_path as "unknown"](https://github.com/ruvnet/claude-flow/issues/1155) | Medium |
 | HK-002 | [#1058 MCP hook handlers are stubs that don't persist data](https://github.com/ruvnet/claude-flow/issues/1058) | High |
 | HK-003 | [#1158 hooks_metrics MCP handler returns hardcoded fake data](https://github.com/ruvnet/claude-flow/issues/1158) | High |
+| HK-004 | [#1175 hooks_session-start ignores daemon.autoStart from settings.json](https://github.com/ruvnet/claude-flow/issues/1175) | High |
 | HW-001 | [#1111 Headless workers hang — stdin pipe never closed](https://github.com/ruvnet/claude-flow/issues/1111) | Critical |
 | HW-002 | [#1112 Headless failures silently swallowed as success](https://github.com/ruvnet/claude-flow/issues/1112) | High |
-| HW-003 | [#1113 Worker scheduling intervals too aggressive](https://github.com/ruvnet/claude-flow/issues/1113) | High |
+| HW-003 | [#1113 Worker scheduling intervals too aggressive + settings ignored](https://github.com/ruvnet/claude-flow/issues/1113) | High |
 | IN-001 | [#1154 intelligence.cjs is a stub that doesn't actually learn](https://github.com/ruvnet/claude-flow/issues/1154) | Critical |
 | MM-001 | [#1152 Remove dead persistPath config option](https://github.com/ruvnet/claude-flow/issues/1152) | Low |
 | NS-001 | [#1123 Discovery ops default to wrong namespace](https://github.com/ruvnet/claude-flow/issues/1123) | Critical |
@@ -178,7 +180,7 @@ Save the returned GitHub issue number for the defect README.md.
 | SG-003 | [#1169 Init missing helpers for --dual, --minimal, hooks, and upgrade paths](https://github.com/ruvnet/claude-flow/issues/1169) | Critical |
 | UI-001 | [#1145 intelligence stats crashes on .toFixed()](https://github.com/ruvnet/claude-flow/issues/1145) | Critical |
 | UI-002 | [#1146 neural status shows "Not loaded"](https://github.com/ruvnet/claude-flow/issues/1146) | Low |
-| DM-006 | [#1114 No log rotation — headless execution logs grow unbounded](https://github.com/ruvnet/claude-flow/issues/1114) | Medium |
+| DM-006 | [#1114 No log rotation — logs grow unbounded](https://github.com/ruvnet/claude-flow/issues/1114) | Medium |
 | HW-004 | [#1117 runWithTimeout rejects but does not kill child process](https://github.com/ruvnet/claude-flow/issues/1117) | Medium |
 <!-- GENERATED:defect-tables:end -->
 
@@ -313,7 +315,7 @@ To target a new file, add a variable to `lib/common.py` following the existing p
 
 ```bash
 # Regenerate all documentation from dynamic discovery
-npm run update-docs
+npm run preflight
 
 # Apply -- should show "Applied: ..."
 bash patch-all.sh --global
@@ -339,7 +341,7 @@ npm test
 - [ ] `patch/{PREFIX}-{NNN}-{slug}/sentinel` created with verification directives
 - [ ] Path variable added to `lib/common.py` (if targeting a new file)
 - [ ] If new category prefix: add one line to `lib/categories.json`
-- [ ] `npm run update-docs` regenerates all doc tables
+- [ ] `npm run preflight` regenerates all doc tables
 - [ ] `bash patch-all.sh` applies successfully
 - [ ] `bash patch-all.sh` is idempotent (0 applied on re-run)
 - [ ] `bash check-patches.sh` shows OK
@@ -355,7 +357,7 @@ Before removing any defect:
 1. Confirm the bug is genuinely fixed upstream or the patch is truly unreachable.
 2. Do NOT remove a defect just because a local workaround exists -- the MCP-level patch may still be needed.
 3. If removing, retire the defect ID permanently. Never reassign a deleted ID to a different GitHub issue.
-4. Run `npm run update-docs` to regenerate all documentation.
+4. Run `npm run preflight` to regenerate all documentation.
 
 ---
 
@@ -450,6 +452,31 @@ Two dependency chains exist:
 
 All other patches are independent.
 
+## Preflight & Pre-Commit Hook
+
+A git pre-commit hook at `hooks/pre-commit` runs automatically on every commit. It calls `npm run preflight:check` (read-only) and `npm test`. If anything is stale or tests fail, the commit is blocked.
+
+**Setup** (one-time, already done for this clone):
+```bash
+git config core.hooksPath hooks
+```
+
+**Before staging**, run:
+```bash
+npm run preflight    # Syncs doc tables, defect counts, version strings, config
+npm test             # Runs all tests
+```
+
+Then `git add -u` to stage the regenerated files.
+
+**What `preflight` syncs**:
+- Defect tables in README.md, CLAUDE.md, npm/README.md (from `patch/*/README.md`)
+- Defect counts in `npm/config.json` (from discovery)
+- `npm/config.json` version.current (from `package.json`)
+- Upstream baseline version in prose (from `npm/config.json` targets)
+
+Manual edits to generated sections (`<!-- GENERATED:*:begin/end -->`) will be overwritten.
+
 ## Key Design Decisions
 
 - **Idempotent**: `patch()` checks if `new` string is already present before replacing.

package/README.md

@@ -1,42 +1,71 @@
 # @sparkleideas/claude-flow-patch
 
-## Contents
-
-- [Quick Start](#quick-start)
-- [CLI Commands](#cli-commands)
-- [How It Works](#how-it-works)
-  - [Sentinel Files](#sentinel-files)
-  - [Target Packages](#target-packages)
-  - [Dependency Order](#dependency-order)
-  - [Key Design Decisions](#key-design-decisions)
-  - [Repository Structure](#repository-structure)
-- [Defect Index](#defect-index)
-- [Init-Script Patches](#init-script-patches)
-- [Compatibility](#compatibility)
-- [Links](#links)
-
----
-
-Community patches for [`@claude-flow/cli`](https://www.npmjs.com/package/@claude-flow/cli) **v3.1.0-alpha.41**, [`ruvector`](https://www.npmjs.com/package/ruvector), and [`ruv-swarm`](https://www.npmjs.com/package/ruv-swarm) **v1.0.20**.
-
-These patches fix 29 defects across 13 categories. They are applied at runtime via idempotent Python scripts that perform targeted string replacements on the npx-cached source files.
-
-<a id="quick-start"></a>
+Patch toolkit for `@claude-flow/cli` runtime and init-script defects.
+
+This package applies verified, idempotent patches to your local CLI source in the npm cache or local `node_modules`, then verifies sentinels and can repair projects initialized before patching.
+
+## What This Solves
+
+`claude-flow init` can generate incomplete or broken project helpers when upstream defects are present. This toolkit addresses known defects in:
+
+- `@claude-flow/cli`
+- `ruvector`
+- `ruv-swarm`
+
+<!-- GENERATED:npm-defects:begin -->
+32 tracked defects across 13 categories.
+
+| Defect | Description | GitHub Issue |
+|--------|-------------|-------------|
+| [CF-001](https://github.com/sparkling/claude-flow-patch/tree/master/patch/010-CF-001-doctor-yaml) | Doctor ignores YAML config files | [#1141](https://github.com/ruvnet/claude-flow/issues/1141) |
+| [CF-002](https://github.com/sparkling/claude-flow-patch/tree/master/patch/020-CF-002-config-export-yaml) | Config export shows hardcoded defaults | [#1142](https://github.com/ruvnet/claude-flow/issues/1142) |
+| [DM-001](https://github.com/sparkling/claude-flow-patch/tree/master/patch/030-DM-001-daemon-log-zero) | daemon.log always 0 bytes | [#1116](https://github.com/ruvnet/claude-flow/issues/1116) |
+| [DM-002](https://github.com/sparkling/claude-flow-patch/tree/master/patch/040-DM-002-cpu-load-threshold) | maxCpuLoad=2.0 blocks all workers on multi-core | [#1138](https://github.com/ruvnet/claude-flow/issues/1138) |
+| [DM-003](https://github.com/sparkling/claude-flow-patch/tree/master/patch/050-DM-003-macos-freemem) | macOS freemem() always ~0% — workers blocked | [#1077](https://github.com/ruvnet/claude-flow/issues/1077) |
+| [DM-004](https://github.com/sparkling/claude-flow-patch/tree/master/patch/060-DM-004-preload-worker-stub) | Preload worker stub + missing from defaults | [#1139](https://github.com/ruvnet/claude-flow/issues/1139) |
+| [DM-005](https://github.com/sparkling/claude-flow-patch/tree/master/patch/070-DM-005-consolidation-worker-stub) | Consolidation worker stub (no decay/rebuild) | [#1140](https://github.com/ruvnet/claude-flow/issues/1140) |
+| [DM-006](https://github.com/sparkling/claude-flow-patch/tree/master/patch/300-DM-006-log-rotation) | No log rotation — logs grow unbounded | [#1114](https://github.com/ruvnet/claude-flow/issues/1114) |
+| [EM-001](https://github.com/sparkling/claude-flow-patch/tree/master/patch/080-EM-001-embedding-ignores-config) | Embedding system ignores project config (model + HNSW dims) | [#1143](https://github.com/ruvnet/claude-flow/issues/1143) |
+| [EM-002](https://github.com/sparkling/claude-flow-patch/tree/master/patch/090-EM-002-transformers-cache-eacces) | @xenova/transformers cache EACCES | [#1144](https://github.com/ruvnet/claude-flow/issues/1144) |
+| [GV-001](https://github.com/sparkling/claude-flow-patch/tree/master/patch/100-GV-001-hnsw-ghost-vectors) | HNSW ghost vectors persist after memory delete | [#1122](https://github.com/ruvnet/claude-flow/issues/1122) |
+| [HK-001](https://github.com/sparkling/claude-flow-patch/tree/master/patch/110-HK-001-post-edit-file-path) | post-edit hook records file_path as "unknown" | [#1155](https://github.com/ruvnet/claude-flow/issues/1155) |
+| [HK-002](https://github.com/sparkling/claude-flow-patch/tree/master/patch/120-HK-002-hooks-tools-stub) | MCP hook handlers are stubs that don't persist data | [#1058](https://github.com/ruvnet/claude-flow/issues/1058) |
+| [HK-003](https://github.com/sparkling/claude-flow-patch/tree/master/patch/130-HK-003-metrics-hardcoded) | hooks_metrics MCP handler returns hardcoded fake data | [#1158](https://github.com/ruvnet/claude-flow/issues/1158) |
+| [HK-004](https://github.com/sparkling/claude-flow-patch/tree/master/patch/135-HK-004-respect-daemon-autostart) | hooks_session-start ignores daemon.autoStart from settings.json | [#1175](https://github.com/ruvnet/claude-flow/issues/1175) |
+| [HW-001](https://github.com/sparkling/claude-flow-patch/tree/master/patch/140-HW-001-stdin-hang) | Headless workers hang — stdin pipe never closed | [#1111](https://github.com/ruvnet/claude-flow/issues/1111) |
+| [HW-002](https://github.com/sparkling/claude-flow-patch/tree/master/patch/150-HW-002-failures-swallowed) | Headless failures silently swallowed as success | [#1112](https://github.com/ruvnet/claude-flow/issues/1112) |
+| [HW-003](https://github.com/sparkling/claude-flow-patch/tree/master/patch/160-HW-003-aggressive-intervals) | Worker scheduling intervals too aggressive + settings ignored | [#1113](https://github.com/ruvnet/claude-flow/issues/1113) |
+| [HW-004](https://github.com/sparkling/claude-flow-patch/tree/master/patch/310-HW-004-runwithtimeout-orphan) | runWithTimeout rejects but does not kill child process | [#1117](https://github.com/ruvnet/claude-flow/issues/1117) |
+| [IN-001](https://github.com/sparkling/claude-flow-patch/tree/master/patch/170-IN-001-intelligence-stub) | intelligence.cjs is a stub that doesn't actually learn | [#1154](https://github.com/ruvnet/claude-flow/issues/1154) |
+| [MM-001](https://github.com/sparkling/claude-flow-patch/tree/master/patch/180-MM-001-memory-persist-path) | Remove dead persistPath config option | [#1152](https://github.com/ruvnet/claude-flow/issues/1152) |
+| [NS-001](https://github.com/sparkling/claude-flow-patch/tree/master/patch/190-NS-001-discovery-default-namespace) | Discovery ops default to wrong namespace | [#1123](https://github.com/ruvnet/claude-flow/issues/1123) |
+| [NS-002](https://github.com/sparkling/claude-flow-patch/tree/master/patch/200-NS-002-targeted-require-namespace) | Store/delete/retrieve fall back to 'default' + accept 'all' | [#581](https://github.com/ruvnet/claude-flow/issues/581) |
+| [NS-003](https://github.com/sparkling/claude-flow-patch/tree/master/patch/210-NS-003-namespace-typo-pattern) | Namespace typo 'pattern' vs 'patterns' | [#1136](https://github.com/ruvnet/claude-flow/issues/1136) |
+| [RS-001](https://github.com/sparkling/claude-flow-patch/tree/master/patch/220-RS-001-better-sqlite3-node24) | ruv-swarm MCP fails on Node 24 — better-sqlite3 missing native bindings | [ruv-FANN#185](https://github.com/ruvnet/ruv-FANN/issues/185) |
+| [RV-001](https://github.com/sparkling/claude-flow-patch/tree/master/patch/230-RV-001-force-learn-tick) | force-learn command calls intel.tick() which doesn't exist | [#1156](https://github.com/ruvnet/claude-flow/issues/1156) |
+| [RV-002](https://github.com/sparkling/claude-flow-patch/tree/master/patch/240-RV-002-trajectory-load) | activeTrajectories not loaded from saved file | [#1157](https://github.com/ruvnet/claude-flow/issues/1157) |
+| [RV-003](https://github.com/sparkling/claude-flow-patch/tree/master/patch/250-RV-003-trajectory-stats-sync) | trajectory-end does not update stats counters | [ruv-FANN#186](https://github.com/ruvnet/ruv-FANN/issues/186) |
+| [SG-001](https://github.com/sparkling/claude-flow-patch/tree/master/patch/260-SG-001-init-settings) | Init generates invalid settings | [#1150](https://github.com/ruvnet/claude-flow/issues/1150) |
+| [SG-003](https://github.com/sparkling/claude-flow-patch/tree/master/patch/270-SG-003-init-helpers-all-paths) | Init missing helpers for --dual, --minimal, hooks, and upgrade paths | [#1169](https://github.com/ruvnet/claude-flow/issues/1169) |
+| [UI-001](https://github.com/sparkling/claude-flow-patch/tree/master/patch/280-UI-001-intelligence-stats-crash) | intelligence stats crashes on .toFixed() | [#1145](https://github.com/ruvnet/claude-flow/issues/1145) |
+| [UI-002](https://github.com/sparkling/claude-flow-patch/tree/master/patch/290-UI-002-neural-status-not-loaded) | neural status shows "Not loaded" | [#1146](https://github.com/ruvnet/claude-flow/issues/1146) |
+<!-- GENERATED:npm-defects:end -->
 
 ## Quick Start
 
-**Patch before init.** Several patches fix the init/generator scripts. If you run `claude-flow init` before patching, the generated `.claude/helpers/` files will be stubs with no learning, no PageRank, and no-op feedback. Always patch first:
+Patch first, then initialize projects.
 
 ```bash
-# 1. Patch first -- fixes the init generators
+# 1) apply patches (global npx cache)
 npx --yes @sparkleideas/claude-flow-patch --global
 
-# 2. Then init (or re-init if already initialized)
-npx @claude-flow/cli@latest init            # fresh project
-npx @claude-flow/cli@latest init upgrade    # existing project
-
-# 3. Verify
+# 2) verify sentinels
 npx --yes @sparkleideas/claude-flow-patch check
+
+# 3) now initialize or upgrade project
+npx @claude-flow/cli@latest init
+# or
+npx @claude-flow/cli@latest init upgrade
 ```
 
 If you already initialized before patching:
@@ -45,19 +74,6 @@ If you already initialized before patching:
 npx --yes @sparkleideas/claude-flow-patch repair --target /path/to/project
 ```
 
-### Target Options
-
-```bash
-npx --yes @sparkleideas/claude-flow-patch                                  # global npx cache (default)
-npx --yes @sparkleideas/claude-flow-patch --global                         # explicit global
-npx --yes @sparkleideas/claude-flow-patch --target ~/my-project            # project's node_modules
-npx --yes @sparkleideas/claude-flow-patch --global --target ~/my-project   # both
-```
-
-`npx @claude-flow/cli` uses local `node_modules` if present, otherwise the global npx cache. Use `--target` to patch a project's local install.
-
-<a id="cli-commands"></a>
-
 ## CLI Commands
 
 | Command | Purpose |
@@ -67,281 +83,46 @@ npx --yes @sparkleideas/claude-flow-patch --global --target ~/my-project   # bot
 | `claude-flow-patch check` | Verify patch sentinels and auto-detect drift |
 | `claude-flow-patch repair --target <dir> [--source auto\|local\|global] [--dry-run]` | Rehydrate `.claude/helpers` in projects initialized before patching |
 
-<a id="how-it-works"></a>
-
-## How It Works
-
-1. `patch-all.sh` locates the `@claude-flow/cli` dist files in the npm/npx cache
-2. Globs `patch/*/fix.py` (numeric prefixes on directories ensure correct execution order)
-3. Concatenates `lib/common.py` with each `fix.py` and runs as a single Python process
-4. Each patch is idempotent: skips if already applied, warns if source changed
-
-The `check-patches.sh` sentinel runs on session start to detect npx cache wipes and auto-reapply. It reads `sentinel` files from each patch directory — no hardcoded patch list.
-
-<a id="sentinel-files"></a>
-
-### Sentinel Files
-
-Each patch directory contains a `sentinel` file that declares how to verify the patch is applied:
-
-```
-grep "pattern to find" relative/path/to/file.js
-absent "pattern that should NOT exist" relative/path.js
-none
-package: ruvector
-```
-
-| Directive | Meaning |
-|-----------|---------|
-| `grep "..." file` | Pass if pattern is found in file (standard check) |
-| `absent "..." file` | Pass if pattern is **not** found (removal check) |
-| `none` | No sentinel — skip verification |
-| `package: X` | Target package (default: `@claude-flow/cli`). Skipped if package not installed |
-
-`check-patches.sh` and `lib/discover.mjs` both read these files dynamically. Adding a new patch requires no edits to any script — just create the `sentinel` file in the new patch directory.
-
-<a id="target-packages"></a>
-
-### Target Packages
-
-| Package | Version | Location | Env var |
-|---------|---------|----------|---------|
-| `@claude-flow/cli` | `3.1.0-alpha.41` | `~/.npm/_npx/*/node_modules/@claude-flow/cli/dist/src/` | `BASE` |
-| `ruvector` | (bundled) | `~/.npm/_npx/*/node_modules/ruvector/bin/cli.js` | `RUVECTOR_CLI` |
-| `ruv-swarm` | `1.0.20` | `~/.npm/_npx/*/node_modules/ruv-swarm/` | (found via glob) |
-
-`BASE` is set by `patch-all.sh`. All path variables in `lib/common.py` derive from it.
-`RUVECTOR_CLI` is set by `patch-all.sh` to the ruvector CLI entry point.
-RS-001 locates its own target via `find`.
-
-<a id="dependency-order"></a>
-
-### Dependency Order
-
-Execution order is controlled by 3-digit numeric prefixes on directory names (e.g. `010-CF-001-*`,
-`170-IN-001-*`). `patch-all.sh` globs `patch/*/fix.py`, which sorts lexicographically — numeric
-prefixes guarantee correct order.
-
-Two dependency chains exist:
-
-| Chain | Directories | Reason |
-|-------|-------------|--------|
-| IN-001 -> SG-003 | `170-IN-001-*` before `270-SG-003-*` | SG-003's patch targets code introduced by IN-001 |
-| NS-001 -> NS-002 -> NS-003 | `190-NS-001-*` before `200-NS-002-*` before `210-NS-003-*` | Sequential namespace fixes |
-
-All other patches are independent.
-
-<a id="key-design-decisions"></a>
-
-### Key Design Decisions
-
-- **Zero-maintenance discovery**: `patch-all.sh`, `check-patches.sh`, and doc generation all discover patches dynamically — no hardcoded lists.
-- **Idempotent**: `patch()` checks if `new` string is already present before replacing.
-- **Non-destructive**: patches only modify the npx cache, never the npm registry package.
-- **Platform-aware**: DM-003 is macOS-only (auto-skipped on Linux).
-- **Sentinel-guarded**: `check-patches.sh` reads `sentinel` files from each patch directory to detect cache wipes and auto-reapply.
-
-<a id="repository-structure"></a>
-
-### Repository Structure
-
-```
-claude-flow-patch/
-  README.md              # This file
-  CLAUDE.md              # Claude Code instructions (defect workflow, policies)
-  AGENTS.md              # Codex agent configuration
-  patch-all.sh           # Apply all patches (globs patch/*/fix.py dynamically)
-  check-patches.sh       # Sentinel: reads patch/*/sentinel files dynamically
-  repair-post-init.sh    # Post-init helper repair
-  lib/
-    common.py            # Shared patch()/patch_all() helpers + path variables
-    discover.mjs         # Dynamic patch discovery — single source of truth
-    categories.json      # Prefix-to-label mapping (e.g. HW → Headless Worker)
-  scripts/
-    update-docs.mjs      # Regenerate doc tables from discover() output
-  patch/
-    {NNN}-{PREFIX}-{NNN}-{slug}/    # NNN = 3-digit execution order
-      README.md          # Defect report: title, severity, root cause, fix
-      fix.py             # Idempotent patch script
-      fix.sh             # Shell-based patch script (EM-002 only)
-      sentinel           # Verification directives for check-patches.sh
-    (29 defect directories total)
-```
-
-<a id="defect-index"></a>
-
-## Defect Index
-
-<!-- GENERATED:defect-index:begin -->
-31 defects across 13 categories.
-
-### CF -- Config & Doctor
-
-| ID | Description <img width="500" height="1" /> | Severity | GitHub&nbsp;Issue |
-|----|-------------|----------|--------------|
-| [CF&#8209;001](patch/010-CF-001-doctor-yaml/) | Doctor ignores YAML config files | Low | [#1141](https://github.com/ruvnet/claude-flow/issues/1141) |
-| [CF&#8209;002](patch/020-CF-002-config-export-yaml/) | Config export shows hardcoded defaults | Medium | [#1142](https://github.com/ruvnet/claude-flow/issues/1142) |
+## Target Behavior
 
-### DM -- Daemon & Workers
-
-| ID | Description <img width="500" height="1" /> | Severity | GitHub&nbsp;Issue |
-|----|-------------|----------|--------------|
-| [DM&#8209;001](patch/030-DM-001-daemon-log-zero/) | daemon.log always 0 bytes | Medium | [#1116](https://github.com/ruvnet/claude-flow/issues/1116) |
-| [DM&#8209;002](patch/040-DM-002-cpu-load-threshold/) | maxCpuLoad=2.0 blocks all workers on multi-core | Critical | [#1138](https://github.com/ruvnet/claude-flow/issues/1138) |
-| [DM&#8209;003](patch/050-DM-003-macos-freemem/) | macOS freemem() always ~0% — workers blocked | Critical | [#1077](https://github.com/ruvnet/claude-flow/issues/1077) |
-| [DM&#8209;004](patch/060-DM-004-preload-worker-stub/) | Preload worker stub + missing from defaults | Enhancement | [#1139](https://github.com/ruvnet/claude-flow/issues/1139) |
-| [DM&#8209;005](patch/070-DM-005-consolidation-worker-stub/) | Consolidation worker stub (no decay/rebuild) | Enhancement | [#1140](https://github.com/ruvnet/claude-flow/issues/1140) |
-| [DM&#8209;006](patch/300-DM-006-log-rotation/) | No log rotation — headless execution logs grow unbounded | Medium | [#1114](https://github.com/ruvnet/claude-flow/issues/1114) |
-
-### EM -- Embeddings & HNSW
-
-| ID | Description <img width="500" height="1" /> | Severity | GitHub&nbsp;Issue |
-|----|-------------|----------|--------------|
-| [EM&#8209;001](patch/080-EM-001-embedding-ignores-config/) | Embedding system ignores project config (model + HNSW dims) | High | [#1143](https://github.com/ruvnet/claude-flow/issues/1143) |
-| [EM&#8209;002](patch/090-EM-002-transformers-cache-eacces/) | @xenova/transformers cache EACCES | Medium | [#1144](https://github.com/ruvnet/claude-flow/issues/1144) |
-
-### GV -- Ghost Vectors
-
-| ID | Description <img width="500" height="1" /> | Severity | GitHub&nbsp;Issue |
-|----|-------------|----------|--------------|
-| [GV&#8209;001](patch/100-GV-001-hnsw-ghost-vectors/) | HNSW ghost vectors persist after memory delete | Medium | [#1122](https://github.com/ruvnet/claude-flow/issues/1122) |
-
-### HK -- Hooks
-
-| ID | Description <img width="500" height="1" /> | Severity | GitHub&nbsp;Issue |
-|----|-------------|----------|--------------|
-| [HK&#8209;001](patch/110-HK-001-post-edit-file-path/) | post-edit hook records file_path as "unknown" | Medium | [#1155](https://github.com/ruvnet/claude-flow/issues/1155) |
-| [HK&#8209;002](patch/120-HK-002-hooks-tools-stub/) | MCP hook handlers are stubs that don't persist data | High | [#1058](https://github.com/ruvnet/claude-flow/issues/1058) |
-| [HK&#8209;003](patch/130-HK-003-metrics-hardcoded/) | hooks_metrics MCP handler returns hardcoded fake data | High | [#1158](https://github.com/ruvnet/claude-flow/issues/1158) |
-
-### HW -- Headless Worker
-
-| ID | Description <img width="500" height="1" /> | Severity | GitHub&nbsp;Issue |
-|----|-------------|----------|--------------|
-| [HW&#8209;001](patch/140-HW-001-stdin-hang/) | Headless workers hang — stdin pipe never closed | Critical | [#1111](https://github.com/ruvnet/claude-flow/issues/1111) |
-| [HW&#8209;002](patch/150-HW-002-failures-swallowed/) | Headless failures silently swallowed as success | High | [#1112](https://github.com/ruvnet/claude-flow/issues/1112) |
-| [HW&#8209;003](patch/160-HW-003-aggressive-intervals/) | Worker scheduling intervals too aggressive | High | [#1113](https://github.com/ruvnet/claude-flow/issues/1113) |
-| [HW&#8209;004](patch/310-HW-004-runwithtimeout-orphan/) | runWithTimeout rejects but does not kill child process | Medium | [#1117](https://github.com/ruvnet/claude-flow/issues/1117) |
-
-### IN -- Intelligence
-
-| ID | Description <img width="500" height="1" /> | Severity | GitHub&nbsp;Issue |
-|----|-------------|----------|--------------|
-| [IN&#8209;001](patch/170-IN-001-intelligence-stub/) | intelligence.cjs is a stub that doesn't actually learn | Critical | [#1154](https://github.com/ruvnet/claude-flow/issues/1154) |
-
-### MM -- Memory Management
-
-| ID | Description <img width="500" height="1" /> | Severity | GitHub&nbsp;Issue |
-|----|-------------|----------|--------------|
-| [MM&#8209;001](patch/180-MM-001-memory-persist-path/) | Remove dead persistPath config option | Low | [#1152](https://github.com/ruvnet/claude-flow/issues/1152) |
-
-### NS -- Memory Namespace
-
-| ID | Description <img width="500" height="1" /> | Severity | GitHub&nbsp;Issue |
-|----|-------------|----------|--------------|
-| [NS&#8209;001](patch/190-NS-001-discovery-default-namespace/) | Discovery ops default to wrong namespace | Critical | [#1123](https://github.com/ruvnet/claude-flow/issues/1123) |
-| [NS&#8209;002](patch/200-NS-002-targeted-require-namespace/) | Store/delete/retrieve fall back to 'default' + accept 'all' | Critical | [#581](https://github.com/ruvnet/claude-flow/issues/581) |
-| [NS&#8209;003](patch/210-NS-003-namespace-typo-pattern/) | Namespace typo 'pattern' vs 'patterns' | Medium | [#1136](https://github.com/ruvnet/claude-flow/issues/1136) |
-
-### RS -- ruv-swarm
-
-| ID | Description <img width="500" height="1" /> | Severity | GitHub&nbsp;Issue |
-|----|-------------|----------|--------------|
-| [RS&#8209;001](patch/220-RS-001-better-sqlite3-node24/) | ruv-swarm MCP fails on Node 24 — better-sqlite3 missing native bindings | Critical | [ruv-FANN#185](https://github.com/ruvnet/ruv-FANN/issues/185) |
-
-### RV -- RuVector Intelligence
-
-| ID | Description <img width="500" height="1" /> | Severity | GitHub&nbsp;Issue |
-|----|-------------|----------|--------------|
-| [RV&#8209;001](patch/230-RV-001-force-learn-tick/) | force-learn command calls intel.tick() which doesn't exist | Medium | [#1156](https://github.com/ruvnet/claude-flow/issues/1156) |
-| [RV&#8209;002](patch/240-RV-002-trajectory-load/) | activeTrajectories not loaded from saved file | High | [#1157](https://github.com/ruvnet/claude-flow/issues/1157) |
-| [RV&#8209;003](patch/250-RV-003-trajectory-stats-sync/) | trajectory-end does not update stats counters | Medium | [ruv-FANN#186](https://github.com/ruvnet/ruv-FANN/issues/186) |
-
-### SG -- Settings Generator
-
-| ID | Description <img width="500" height="1" /> | Severity | GitHub&nbsp;Issue |
-|----|-------------|----------|--------------|
-| [SG&#8209;001](patch/260-SG-001-init-settings/) | Init generates invalid settings | High | [#1150](https://github.com/ruvnet/claude-flow/issues/1150) |
-| [SG&#8209;003](patch/270-SG-003-init-helpers-all-paths/) | Init missing helpers for --dual, --minimal, hooks, and upgrade paths | Critical | [#1169](https://github.com/ruvnet/claude-flow/issues/1169) |
-
-### UI -- Display & Cosmetic
-
-| ID | Description <img width="500" height="1" /> | Severity | GitHub&nbsp;Issue |
-|----|-------------|----------|--------------|
-| [UI&#8209;001](patch/280-UI-001-intelligence-stats-crash/) | intelligence stats crashes on .toFixed() | Critical | [#1145](https://github.com/ruvnet/claude-flow/issues/1145) |
-| [UI&#8209;002](patch/290-UI-002-neural-status-not-loaded/) | neural status shows "Not loaded" | Low | [#1146](https://github.com/ruvnet/claude-flow/issues/1146) |
-<!-- GENERATED:defect-index:end -->
-
-<a id="init-script-patches"></a>
-
-## Init-Script Patches (Local Project Action Required)
-
-Five patches target the **init/generator scripts** (`executor.js`, `settings-generator.js`, `helpers-generator.js`). These fix the code that *generates* your `.claude/` project files -- but applying patches does **not** update files already generated in your project. You must take one additional step.
-
-### Affected Defects
-
-| ID | Generator patched | Local file affected | Problem if not refreshed |
-|----|-------------------|---------------------|--------------------------|
-| IN-001 | `init/executor.js` | `.claude/helpers/intelligence.cjs` | 197-line stub: no PageRank, no graph, `feedback()` is a no-op, no learning |
-| HK-001 | `init/helpers-generator.js` | `.claude/helpers/hook-handler.cjs` | Reads env vars instead of stdin JSON; post-edit logs `file: "unknown"` |
-| SG-001 | `init/settings-generator.js` | `.claude/settings.json` | May contain invalid hook events, broad permissions, relative paths |
-| SG-002 | `init/executor.js` | `.claude/helpers/*.js` / `*.cjs` | Missing .js/.cjs compat copies; `hook-handler.cjs` require() calls fail silently |
-| MM-001 | `init/executor.js` | `.claude-flow/config.yaml` | Misleading `persistPath` setting that nothing reads |
-
-### How to Fix
-
-**Option A: Run `repair`** (recommended)
-
-```bash
-npx --yes @sparkleideas/claude-flow-patch --global
-npx --yes @sparkleideas/claude-flow-patch repair --target .
-npx --yes @sparkleideas/claude-flow-patch apply SG-002       # apply a single patch
-```
-
-This copies patched helper files into your project and creates any missing .js/.cjs compat copies.
+| Flag | Patched location |
+|---|---|
+| (none) | Global npx cache (default) |
+| `--global` | `~/.npm/_npx/*/node_modules/...` |
+| `--target <dir>` | `<dir>/node_modules/...` |
+| `--global --target <dir>` | Both locations |
 
-**Option B: Copy full helpers from the package manually**
+Use `--target` when your project has a local `@claude-flow/cli` install.
 
-```bash
-npx --yes @sparkleideas/claude-flow-patch --global
-SRC=$(find ~/.npm/_npx -path '*/@claude-flow/cli/.claude/helpers' -type d 2>/dev/null | head -1)
-for f in intelligence.cjs hook-handler.cjs session.js learning-service.mjs metrics-db.mjs statusline.cjs; do
-  [ -f "$SRC/$f" ] && cp "$SRC/$f" .claude/helpers/ && echo "Copied: $f"
-done
-```
+## Why `repair` Exists
 
-**Option C: Re-run init upgrade** (regenerates from patched scripts)
+Some fixes patch generator code (`init/executor.js`, `init/settings-generator.js`, `init/helpers-generator.js`). Those fixes affect future generated files only.
 
-```bash
-npx --yes @sparkleideas/claude-flow-patch --global
-npx @claude-flow/cli@latest init upgrade --force
-```
-
-Caution: Option C may overwrite other customizations in `.claude/`.
+If your repo was initialized before patching, existing `.claude/helpers/*` can remain stale. `repair` copies patched helper files into the target repo and creates .js/.cjs compatibility copies where needed.
 
-### Why This Happens
+## How Patching Works
 
-These patches fix the **generator functions** inside the npm package (e.g., `generateIntelligenceStub()` in `executor.js`). When the generator runs via `claude-flow init`, it produces the project files in `.claude/helpers/`. If your project was initialized *before* patches were applied, the stubs are already on disk. Patches only modify the npm package source -- they do not touch files already generated in your project.
+1. Locate CLI/package targets.
+2. Concatenate shared helpers (`lib/common.py`) with each defect `fix.py`.
+3. Apply string-based replacements with idempotency guards.
+4. Validate with sentinel checks.
 
-Additionally, `init upgrade` only force-overwrites 3 "critical" helpers (`auto-memory-hook.mjs`, `hook-handler.cjs`, `intelligence.cjs`). The other 30+ helper files (shell scripts for daemon management, health monitoring, security scanning, swarm hooks, etc.) are only copied on fresh `init`, not on upgrade. If these are missing, use Option A above.
+Design guarantees:
 
-<a id="compatibility"></a>
+- **Idempotent**: safe to re-run
+- **Non-destructive**: modifies local cache/install only
+- **Ordered**: dependent patches applied in sequence (e.g. NS-001 -> NS-002 -> NS-003)
+- **Platform-aware**: macOS-specific patches auto-skip on Linux
 
 ## Compatibility
 
-- Tested against `@claude-flow/cli@3.1.0-alpha.41` and `ruv-swarm@1.0.20`
-- Requires Python 3.6+ and Bash
-- Works on Linux and macOS (DM-003 is macOS-only, auto-skipped on Linux)
-
-<a id="links"></a>
+- Tested with `@claude-flow/cli@3.1.0-alpha.41`
+- Tested with `ruv-swarm@1.0.20`
+- Requires Bash + Python 3.6+
+- Linux and macOS supported
 
 ## Links
 
 - Homepage: https://sparklingideas.co.uk/claude-flow/patch
-- Package: https://www.npmjs.com/package/@sparkleideas/claude-flow-patch
 - GitHub: https://github.com/sparkling/claude-flow-patch
 - Issues: https://github.com/sparkling/claude-flow-patch/issues
-
-## License
-
-MIT

package/check-patches.sh

@@ -165,10 +165,14 @@ if [ -x "$SCRIPT_DIR/patch-all.sh" ]; then
   if [[ -n "$TARGET_DIR" ]]; then REAPPLY_ARGS+=(--target "$TARGET_DIR"); fi
   bash "$SCRIPT_DIR/patch-all.sh" "${REAPPLY_ARGS[@]}"
   echo ""
-  echo "[PATCHES] Auto-reapplied. Restarting daemon..."
+  echo "[PATCHES] Auto-reapplied. Stopping existing daemons..."
   npx @claude-flow/cli@latest daemon stop 2>/dev/null
+  # Fallback: kill by PID file if daemon stop missed an orphan (project-scoped, not global)
+  _pid=$(cat .claude-flow/daemon.pid 2>/dev/null)
+  if [ -n "$_pid" ]; then kill "$_pid" 2>/dev/null || true; rm -f .claude-flow/daemon.pid; fi
+  sleep 1
   npx @claude-flow/cli@latest daemon start 2>/dev/null
-  echo "[PATCHES] Daemon restarted with patched code."
+  echo "[PATCHES] Daemon restarted in background (PID: $(cat .claude-flow/daemon.pid 2>/dev/null || echo 'unknown'))"
   echo ""
 else
   echo "[PATCHES] ERROR: patch-all.sh not found at $SCRIPT_DIR"

package/package.json

@@ -1,15 +1,16 @@
 {
   "name": "@sparkleideas/claude-flow-patch",
-  "version": "3.1.0-alpha.44.patch.4",
+  "version": "3.1.0-alpha.44.patch.6",
   "description": "Patch toolkit for @claude-flow/cli init/runtime defects with verify and post-init repair commands",
   "scripts": {
+    "preflight": "node scripts/preflight.mjs",
+    "preflight:check": "node scripts/preflight.mjs --check",
+    "prepublishOnly": "npm run preflight && npm test",
     "test": "node --test tests/*.test.mjs",
     "package": "bash npm/package.sh",
     "package:dry": "bash npm/package.sh --dry-run",
     "publish:npm": "bash npm/publish.sh",
     "publish:dry": "bash npm/publish.sh --dry-run",
-    "update-docs": "node scripts/update-docs.mjs",
-    "update-docs:check": "node scripts/update-docs.mjs --check",
     "upstream-log": "node scripts/upstream-log.mjs",
     "guidance:analyze": "guidance analyze",
     "guidance:optimize": "guidance autopilot --once --apply --source manual",

package/patch/135-HK-004-respect-daemon-autostart/README.md

@@ -0,0 +1,11 @@
+# HK-004: hooks_session-start ignores daemon.autoStart from settings.json
+**Severity**: High
+**GitHub**: [#1175](https://github.com/ruvnet/claude-flow/issues/1175)
+## Root Cause
+`hooks_session-start` handler in `hooks-tools.js` (line ~1216) determines daemon auto-start solely from the MCP call parameter: `const shouldStartDaemon = params.startDaemon !== false;`. It never reads `claudeFlow.daemon.autoStart` from `.claude/settings.json`. Setting `autoStart: false` in settings.json has no effect — the daemon always starts on session-start.
+## Fix
+Wrap the `shouldStartDaemon` assignment in an IIFE that first checks the MCP parameter, then reads settings.json. If `claudeFlow.daemon.autoStart === false`, returns false. Falls back to true on any read/parse error.
+## Files Patched
+- mcp-tools/hooks-tools.js
+## Ops
+1 op in fix.py

package/patch/135-HK-004-respect-daemon-autostart/fix.py

@@ -0,0 +1,14 @@
+# HK-004: hooks_session-start ignores daemon.autoStart from settings.json
+# GitHub: #1175
+patch("HK-004: respect daemon autoStart setting",
+    MCP_HOOKS,
+    "        const shouldStartDaemon = params.startDaemon !== false;",
+    """        const shouldStartDaemon = (() => {
+            if (params.startDaemon === false) return false;
+            try {
+                const sp = join(process.cwd(), '.claude', 'settings.json');
+                const s = JSON.parse(readFileSync(sp, 'utf-8'));
+                if (s?.claudeFlow?.daemon?.autoStart === false) return false;
+            } catch { /* no settings or unreadable — default to true */ }
+            return true;
+        })();""")

package/patch/135-HK-004-respect-daemon-autostart/sentinel

@@ -0,0 +1 @@
+grep "claudeFlow?.daemon?.autoStart" mcp-tools/hooks-tools.js

package/patch/160-HW-003-aggressive-intervals/README.md

@@ -1,11 +1,11 @@
-# HW-003: Worker scheduling intervals too aggressive
+# HW-003: Worker scheduling intervals too aggressive + settings ignored
 **Severity**: High
 **GitHub**: [#1113](https://github.com/ruvnet/claude-flow/issues/1113)
 ## Root Cause
-`DEFAULT_WORKERS` uses pre-headless intervals (audit: 10m, optimize: 15m, testgaps: 20m). ADR-020 specifies longer intervals (30/60/60m) for headless workers that invoke Claude.
+`DEFAULT_WORKERS` uses pre-headless intervals (audit: 10m, optimize: 15m, testgaps: 20m). ADR-020 specifies longer intervals (30/60/60m) for headless workers that invoke Claude. Additionally, `daemon.schedules` from `.claude/settings.json` is never read — user-configured intervals are completely ignored.
 ## Fix
-Align intervals to ADR-020: audit 30m, optimize 60m, testgaps 60m.
+(A) Align hardcoded intervals to ADR-020: audit 30m, optimize 60m, testgaps 60m. (B) After setting default workers, read `claudeFlow.daemon.schedules` from `.claude/settings.json` and merge user-configured intervals/enabled flags into matching workers. Supports string formats ("1h", "30m", "10s") and raw milliseconds.
 ## Files Patched
 - services/worker-daemon.js
 ## Ops
-3 ops in fix.py
+4 ops in fix.py

package/patch/160-HW-003-aggressive-intervals/fix.py

@@ -14,3 +14,39 @@ patch("3: testgaps 60m",
     WD,
     "type: 'testgaps', intervalMs: 20 * 60 * 1000",
     "type: 'testgaps', intervalMs: 60 * 60 * 1000")
+
+# HW-003 extension: Read daemon.schedules from .claude/settings.json
+# and merge user-configured intervals into worker defaults
+patch("3: settings-driven intervals",
+    WD,
+    "            workers: config?.workers ?? DEFAULT_WORKERS,",
+    """            workers: (() => {
+                const base = config?.workers ?? DEFAULT_WORKERS;
+                try {
+                    const sp = join(projectRoot, '.claude', 'settings.json');
+                    const s = JSON.parse(readFileSync(sp, 'utf-8'));
+                    const schedules = s?.claudeFlow?.daemon?.schedules;
+                    if (!schedules || typeof schedules !== 'object') return base;
+                    const parseInterval = (v) => {
+                        if (typeof v === 'number') return v;
+                        if (typeof v !== 'string') return null;
+                        const m = v.match(/^(\\d+(?:\\.\\d+)?)\\s*(ms|s|m|h)$/i);
+                        if (!m) return null;
+                        const n = parseFloat(m[1]);
+                        switch (m[2].toLowerCase()) {
+                            case 'ms': return n;
+                            case 's': return n * 1000;
+                            case 'm': return n * 60 * 1000;
+                            case 'h': return n * 3600 * 1000;
+                            default: return null;
+                        }
+                    };
+                    return base.map(w => {
+                        const sched = schedules[w.type];
+                        if (!sched) return w;
+                        const iv = parseInterval(sched.interval ?? sched.intervalMs);
+                        const en = typeof sched.enabled === 'boolean' ? sched.enabled : w.enabled;
+                        return { ...w, ...(iv !== null ? { intervalMs: iv } : {}), enabled: en };
+                    });
+                } catch { return base; }
+            })(),""")

package/patch/160-HW-003-aggressive-intervals/sentinel

@@ -1 +1,3 @@
-grep "30 * 60 * 1000" services/worker-daemon.js
+grep "intervalMs: 30" services/worker-daemon.js
+grep "parseInterval" services/worker-daemon.js
+grep "daemon?.schedules" services/worker-daemon.js

package/patch/170-IN-001-intelligence-stub/fix.py

@@ -10,15 +10,12 @@
 # 2 ops: upgrade path (line ~398) + writeHelpers path (line ~943)
 
 # Op 1: Upgrade path — when findSourceHelpersDir() returns null
+# old/new strings stop BEFORE generatedCritical to avoid overlap with SG-003j
 patch("IN-001a: upgrade fallback reads real intelligence.cjs",
     EXECUTOR,
     """        else {
             // Source not found (npx with broken paths) — use generated fallbacks
-            const generatedCritical = {
-                'hook-handler.cjs': generateHookHandler(),
-                'intelligence.cjs': generateIntelligenceStub(),
-                'auto-memory-hook.mjs': generateAutoMemoryHook(),
-            };""",
+            const generatedCritical = {""",
     """        else {
             // Source not found (npx with broken paths) — use generated fallbacks
             // IN-001: Try reading real intelligence.cjs from package before using stub
@@ -29,11 +26,7 @@ patch("IN-001a: upgrade fallback reads real intelligence.cjs",
                     intelligenceContent = fs.readFileSync(realPath, 'utf-8');
                 }
             } catch { /* use stub */ }
-            const generatedCritical = {
-                'hook-handler.cjs': generateHookHandler(),
-                'intelligence.cjs': intelligenceContent,
-                'auto-memory-hook.mjs': generateAutoMemoryHook(),
-            };""")
+            const generatedCritical = {""")
 
 # Op 2: writeHelpers path — fresh init when source dir not found
 patch("IN-001b: writeHelpers fallback reads real intelligence.cjs",

package/patch/270-SG-003-init-helpers-all-paths/fix.py

@@ -122,11 +122,12 @@ patch_all("SG-003i: source hook-handler requires memory.cjs",
 
 # Op 6: Fix executeUpgrade() fallback — generatedCritical missing router/session/memory
 # When source helpers aren't found, only 3 files were generated but hook-handler.cjs needs 6
+# Also replaces generateIntelligenceStub() → intelligenceContent (set by IN-001a above)
 patch("SG-003j: upgrade fallback generates router/session/memory",
     EXECUTOR,
     """            const generatedCritical = {
                 'hook-handler.cjs': generateHookHandler(),
-                'intelligence.cjs': intelligenceContent,
+                'intelligence.cjs': generateIntelligenceStub(),
                 'auto-memory-hook.mjs': generateAutoMemoryHook(),
             };""",
     """            const generatedCritical = {

package/patch/300-DM-006-log-rotation/README.md

@@ -1,11 +1,12 @@
-# DM-006: No log rotation — headless execution logs grow unbounded
+# DM-006: No log rotation — logs grow unbounded
 **Severity**: Medium
 **GitHub**: [#1114](https://github.com/ruvnet/claude-flow/issues/1114)
 ## Root Cause
-`logExecution()` in `headless-worker-executor.js` creates 2-3 log files per worker run (~75 KB each) but has zero cleanup. No rotation, no max file count, no TTL. At current daemon intervals this accumulates ~23 MB/day, ~702 MB/month.
+(Headless) `logExecution()` in `headless-worker-executor.js` creates 2-3 log files per worker run (~75 KB each) but has zero cleanup. No rotation, no max file count, no TTL. At current daemon intervals this accumulates ~23 MB/day, ~702 MB/month. (Main) `startBackgroundDaemon()` in `daemon.js` opens `daemon.log` in append mode and never truncates. A single long-running daemon can grow daemon.log to 100+ GB.
 ## Fix
-(A) Add `unlinkSync` and `statSync` to the ESM import. (B) Call `cleanupOldLogs()` from `ensureLogDir()` so cleanup runs on each execution cycle. (C) Add `cleanupOldLogs()` method: removes `.log` files older than 7 days or beyond a 500-file cap, keeping newest files.
+(A) Add `unlinkSync` and `statSync` to the headless executor ESM import. (B) Call `cleanupOldLogs()` from `ensureLogDir()` so cleanup runs on each headless execution cycle. (C) Add `cleanupOldLogs()` method: removes `.log` files older than 7 days or beyond a 500-file cap, keeping newest files. (D) In `daemon.js`, before opening `daemon.log` for append, check its size with `fs.statSync()` and rotate to `daemon.log.1` if > 50MB.
 ## Files Patched
 - services/headless-worker-executor.js
+- commands/daemon.js
 ## Ops
-3 ops in fix.py
+4 ops in fix.py

package/patch/300-DM-006-log-rotation/fix.py

@@ -56,3 +56,17 @@ patch("DM-006c: add cleanupOldLogs method",
         catch { /* ignore cleanup errors */ }
     }
     logExecution(executionId, type, content) {""")
+
+# DM-006 extension: Main daemon.log rotation in daemon.js
+# Before opening daemon.log for append, check size and rotate if > 50MB
+patch("DM-006d: main daemon.log rotation",
+    DJ,
+    "    const logFile = join(logsDir, 'daemon.log');",
+    """    const logFile = join(logsDir, 'daemon.log');
+    // Rotate main daemon.log if > 50MB
+    try {
+        const logStat = fs.statSync(logFile);
+        if (logStat.size > 50 * 1024 * 1024) {
+            fs.renameSync(logFile, logFile + '.1');
+        }
+    } catch { /* file doesn't exist yet or stat failed — ignore */ }""")

package/patch/300-DM-006-log-rotation/sentinel

@@ -1 +1,2 @@
 grep "cleanupOldLogs" services/headless-worker-executor.js
+grep "Rotate main daemon.log" commands/daemon.js

package/scripts/{update-docs.mjs → preflight.mjs}

@@ -1,9 +1,10 @@
 #!/usr/bin/env node
-// scripts/update-docs.mjs — Regenerate documentation from dynamic patch discovery.
-// Updates: README.md, CLAUDE.md, npm/README.md, npm/config.json
+// scripts/preflight.mjs — Pre-commit/pre-publish consistency check.
+// Syncs: doc tables, defect counts, version strings across all files.
+// Source of truth: package.json (version), npm/config.json (targets), patch/*/ (defects).
 //
-// Usage: node scripts/update-docs.mjs [--check]
-//   --check  Exit 1 if docs are out of date (for CI), don't write.
+// Usage: node scripts/preflight.mjs [--check]
+//   --check  Exit 1 if anything is out of date (for hooks/CI), don't write.
 
 import { readFileSync, writeFileSync } from 'node:fs';
 import { resolve, dirname } from 'node:path';
@@ -17,6 +18,14 @@ const checkOnly = process.argv.includes('--check');
 const data = discover();
 const { patches, categories, stats } = data;
 
+// ── Sources of truth ──
+
+const pkgJson = JSON.parse(readFileSync(resolve(ROOT, 'package.json'), 'utf-8'));
+const configJson = JSON.parse(readFileSync(resolve(ROOT, 'npm', 'config.json'), 'utf-8'));
+const pkgVersion = pkgJson.version;
+const cliTarget = configJson.targets['@claude-flow/cli'];
+const swarmTarget = configJson.targets['ruv-swarm'];
+
 // ── Helpers ──
 
 /** Group patches by prefix, preserving sort order. */
@@ -57,6 +66,20 @@ function replaceMarkerSection(filePath, markerName, newContent) {
   return true;
 }
 
+/**
+ * Replace all occurrences of a version string in a file.
+ * Returns true if content changed.
+ */
+function syncVersionInFile(filePath, oldVersion, newVersion, label) {
+  if (oldVersion === newVersion) return false;
+  const text = readFileSync(filePath, 'utf-8');
+  if (!text.includes(oldVersion)) return false;
+  const updated = text.replaceAll(oldVersion, newVersion);
+  if (updated === text) return false;
+  if (!checkOnly) writeFileSync(filePath, updated);
+  return true;
+}
+
 // ── Generate README.md defect index ──
 
 function generateReadmeIndex() {
@@ -142,13 +165,21 @@ function updateNpmReadme() {
   );
 }
 
-// ── Update npm/config.json counts ──
+// ── Sync npm/config.json (version + defect counts) ──
 
 function updateNpmConfig() {
   const filePath = resolve(ROOT, 'npm', 'config.json');
   const config = JSON.parse(readFileSync(filePath, 'utf-8'));
 
   let changed = false;
+
+  // Sync version.current from package.json
+  if (config.version?.current !== pkgVersion) {
+    config.version.current = pkgVersion;
+    changed = true;
+  }
+
+  // Sync defect counts from discovery
   if (config.defects?.total !== stats.total) {
     config.defects.total = stats.total;
     changed = true;
@@ -168,41 +199,51 @@ function updateNpmConfig() {
 
 let anyChanged = false;
 
-const readmeChanged = replaceMarkerSection(
-  resolve(ROOT, 'README.md'),
-  'defect-index',
-  generateReadmeIndex()
-);
-if (readmeChanged) {
+function report(changed, label) {
+  if (!changed) return;
   anyChanged = true;
-  console.log(checkOnly ? 'STALE: README.md' : 'Updated: README.md');
+  console.log(checkOnly ? `STALE: ${label}` : `Updated: ${label}`);
 }
 
-const claudeChanged = replaceMarkerSection(
-  resolve(ROOT, 'CLAUDE.md'),
-  'defect-tables',
-  generateClaudeTables()
+// 1. Doc tables (marker-based sections)
+report(
+  replaceMarkerSection(resolve(ROOT, 'README.md'), 'defect-index', generateReadmeIndex()),
+  'README.md (defect index)'
 );
-if (claudeChanged) {
-  anyChanged = true;
-  console.log(checkOnly ? 'STALE: CLAUDE.md' : 'Updated: CLAUDE.md');
-}
-
-const npmReadmeChanged = updateNpmReadme();
-if (npmReadmeChanged) {
-  anyChanged = true;
-  console.log(checkOnly ? 'STALE: npm/README.md' : 'Updated: npm/README.md');
-}
-
-const npmConfigChanged = updateNpmConfig();
-if (npmConfigChanged) {
-  anyChanged = true;
-  console.log(checkOnly ? 'STALE: npm/config.json' : 'Updated: npm/config.json');
+report(
+  replaceMarkerSection(resolve(ROOT, 'CLAUDE.md'), 'defect-tables', generateClaudeTables()),
+  'CLAUDE.md (defect tables)'
+);
+report(updateNpmReadme(), 'npm/README.md (defect list)');
+
+// 2. Config sync (version + counts)
+report(updateNpmConfig(), 'npm/config.json (version/counts)');
+
+// 3. Upstream baseline version in prose (sync from npm/config.json targets)
+// Find any stale version strings and replace with current targets.
+// We scan for the pattern v?X.Y.Z-alpha.N and replace if it doesn't match config.
+const versionFiles = ['README.md', 'CLAUDE.md', 'npm/README.md', 'AGENTS.md'];
+for (const file of versionFiles) {
+  const filePath = resolve(ROOT, file);
+  let text;
+  try { text = readFileSync(filePath, 'utf-8'); } catch { continue; }
+
+  let updated = text;
+
+  // Sync @claude-flow/cli version references
+  // Match patterns like **v3.1.0-alpha.NN** or `3.1.0-alpha.NN` or @3.1.0-alpha.NN
+  const cliRe = /(?<=[@`*v])3\.1\.0-alpha\.\d+/g;
+  updated = updated.replace(cliRe, cliTarget);
+
+  if (updated !== text) {
+    if (!checkOnly) writeFileSync(filePath, updated);
+    report(true, `${file} (upstream baseline)`);
+  }
 }
 
 if (!anyChanged) {
-  console.log('All docs are up to date.');
+  console.log('All files are up to date.');
 } else if (checkOnly) {
-  console.log('\nDocs are out of date. Run: npm run update-docs');
+  console.log('\nFiles are out of date. Run: npm run preflight');
   process.exit(1);
 }