@softspark/ai-toolkit 4.12.0 → 4.14.0

package/CHANGELOG.md

@@ -7,6 +7,46 @@ Versioning follows [Semantic Versioning](https://semver.org/).
 
 ---
 
+## v4.14.0 — Model refresh + effort-aware routing (2026-07-14)
+
+Minor release. Refreshes Claude model IDs to the current generation across the toolkit, teaches the routing skill to route on `effort` (not just model tier), and moves two pure-retrieval agents to Haiku. No skill/agent/hook count change (44 agents, 108 skills, 1216 tests).
+
+### Changed
+- **Canonical model IDs bumped** — `scripts/_common.py` `DEFAULT_CLAUDE_MODELS` now resolves `opus → claude-opus-4-8` and `sonnet → claude-sonnet-5` (`haiku → claude-haiku-4-5` unchanged). This is the single source of truth generators emit.
+- **`model-routing-patterns` skill refreshed** — tier table updated to Haiku 4.5 / Sonnet 5 / Opus 4.8 with per-1M pricing, added a **Fable 5** row plus a "not the default best model" caveat (target for "strongest model" stays `claude-opus-4-8`), and a new **Effort routing** section (effort is the cheaper lever before swapping models — it does not invalidate the prompt cache). Fallback-chain example IDs updated.
+- **Stale example model IDs refreshed** — `json-mode-patterns`, `prompt-caching-patterns`, and `skill-creator` code samples now reference `claude-opus-4-8`.
+- **Pure-retrieval agents routed to Haiku** — `explorer-agent` (Read/Grep/Glob only) and `fact-checker` (Read only) changed from `model: sonnet` to `model: haiku`, closing the gap where the routing skill recommended Haiku workers but no agent used the tier.
+
+## v4.13.0 — Claude app delivery + Devin sunset cleanup (2026-07-10)
+
+Minor release. Adds a validated distribution path for Claude Chat/Desktop/Cowork, fixes the core plugin contract, and removes the retired Cascade hook surface. Test count: 1208 → 1216 (seven deprecated Cascade assertions removed; six Claude app contract tests, three editor migration/propagation tests, two state-isolation/cleanup tests, two untagged-hook diagnostics, and two dry-run regressions added).
+
+### Added
+- **Claude app plugin export** — `ai-toolkit claude-app export [--output FILE] [--no-custom-rules] [--verify]` creates a deterministic plugin ZIP for `Customize > Plugins` and a sibling Cowork global-instructions file. The archive bundles skills, agents, app-relative hooks, required helper scripts, core rules, and registered user rules (opt-out for shareable bundles).
+- **App-native rules delivery** — common and standalone rule sources are rendered into the `ai-toolkit-rules` skill because Claude Chat/Cowork does not read Claude Code's `CLAUDE.md` or `~/.claude/rules/` surfaces.
+- **Claude app ecosystem target** — registry/tooling now tracks 13 targets: Claude Code, Claude Chat/Cowork, and 11 editor integrations.
+
+### Fixed
+- **Official plugin manifest** — changed `repository` from an invalid object to the required URL string, removed four ignored non-schema fields, and declared the generated app skill/hook component paths. `claude plugin validate` now passes on a clean staged plugin.
+- **Cowork hook portability** — generated plugin hooks use `${CLAUDE_PLUGIN_ROOT}` and bundle their Python helpers instead of assuming a separate `~/.softspark/ai-toolkit/hooks` install.
+- **Untagged hook diagnostics** — `doctor` and `config-desync-guard.sh` now recognize canonical hook behavior when Claude rewrites settings without preserving ai-toolkit's private `_source` marker, eliminating false "hooks missing" warnings.
+- **Update dry-run safety** — `ai-toolkit update --dry-run` and `--list` no longer invoke registered-project updates or create `projects.lock` after the read-only preview.
+- **Cline dry-run accuracy** — global preview now reports reuse of Cline-compatible `~/.claude/skills` instead of claiming it will create a redundant `~/.cline/skills` pointer.
+- **Global/local state isolation** — project-local installs and registered-project propagation no longer overwrite global `state.json` modules/profile with the last project's detected languages; project metadata remains in `projects.json`.
+- **Detected-language state cleanup** — explicit module installs now clear stale `auto_detected_languages` metadata instead of reporting an obsolete last-detected language in `status`.
+- **Registered-project editor/profile propagation** — explicit `ai-toolkit update --editors/--profile ...` overrides now reach registered projects; without an override, each project's saved editor set and profile are preserved.
+- **Devin skill discovery** — emit a single canonical pointer under `.windsurf/skills` instead of also writing `.devin/skills`. Both are in Devin's documented eight-path discovery list (all scanned in every repo), so one pointer keeps skills discoverable while avoiding duplicate registration.
+
+### Removed
+- **Cascade hook generator** — removed `generate_windsurf_hooks.py`, `.windsurf/hooks.json` install wiring, tests, and docs after Cascade's 2026-07-01 sunset. `generate_devin_hooks.py` / `.devin/hooks.v1.json` is the sole Windsurf-family hook surface.
+- **Retired path migration** — local Windsurf/Devin updates strip only ai-toolkit-owned entries from legacy `.windsurf/hooks.json` and remove the managed `.devin/skills/ai-toolkit-skill-catalogue` pointer while preserving user content.
+
+### Ecosystem
+- **Class B/F — Claude app plugins**: adopted the official plugin surface shared by Chat/Desktop/Cowork; skills work across Chat and Cowork while hooks/sub-agents are Cowork-only.
+- **Class D — Devin/Cascade**: completed the scheduled removal and refreshed current skill paths. Devin Desktop v3.4.27 and its v3.4.22 skill-permission change were reviewed; no permission field is needed for ai-toolkit's non-executing pointer skill.
+- **Class A — editor docs**: reviewed content-only drift for Cursor, Copilot, Gemini, Cline, Augment, and opencode; no generator contract changes were found.
+- **Class A — versions/docs**: refreshed Claude Code 2.1.206 and Codex CLI 0.144.1 after reviewing their official releases/current config surfaces.
+
 ## v4.12.0 — Global-first editor installs + Codex path fix (2026-07-02)
 
 Minor release. Expands global (HOME-scoped) installs to every editor with a documented, merge-safe file surface, and fixes a Codex global-install bug where instructions landed in a file Codex never reads. Test count: 1203 → 1208.

package/README.md

@@ -1,22 +1,21 @@
 # ai-toolkit
 
-> Professional-grade AI coding toolkit with multi-platform support. Machine-enforced safety, 108 skills, 44 agents, expanded lifecycle hooks, persona presets, experimental opt-in plugin packs, and benchmark tooling — works with Claude, Cursor, Windsurf, Copilot, Gemini, Cline, Roo Code, Aider, Augment, Google Antigravity, Codex CLI, and opencode, ready in 60 seconds.
+> Professional-grade AI coding toolkit with multi-platform support. Machine-enforced safety, 108 skills, 44 agents, expanded lifecycle hooks, persona presets, experimental opt-in plugin packs, and benchmark tooling — works with Claude Code, Claude Chat/Cowork, Cursor, Devin, Copilot, Gemini, Cline, Roo/Zoo Code, Aider, Augment, Google Antigravity, Codex CLI, and opencode.
 
 [![CI](https://github.com/softspark/ai-toolkit/actions/workflows/ci.yml/badge.svg)](https://github.com/softspark/ai-toolkit/actions/workflows/ci.yml)
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](LICENSE)
 [![Skills](https://img.shields.io/badge/skills-108-brightgreen)](app/skills/)
 [![Agents](https://img.shields.io/badge/agents-44-blue)](app/agents/)
-[![Tests](https://img.shields.io/badge/tests-1208%20passing-success)](tests/)
+[![Tests](https://img.shields.io/badge/tests-1216%20passing-success)](tests/)
 
-## What's New in v4.12.0
+## What's New in v4.14.0
 
-v4.12.0 makes global (HOME-scoped) installs first-class for every editor with a documented file surface, and fixes a Codex global-install bug.
+v4.14.0 refreshes Claude model IDs to the current generation and makes model routing effort-aware.
 
-- **Codex global path fix**: `install --editors codex` now writes `~/.codex/AGENTS.md` (the file Codex actually reads) instead of the never-loaded `~/AGENTS.md`; upgrades strip the stale section.
-- **Global installs for 7 editors**: Gemini (hooks/commands/skills), Augment (agents/commands/hooks), Roo/Zoo (skills), Windsurf (`~/.config/devin/AGENTS.md`), plus new global support for Cursor (`~/.cursor/hooks.json`), GitHub Copilot (`~/.copilot/`), and Antigravity (skill pointer).
-- **Roo → Zoo Code**: retargeted the dead `RooCodeInc/Roo-Code` release feed to the active successor `Zoo-Code-Org/Zoo-Code`.
-- **Ecosystem refresh**: Gemini v0.49.0, Codex 0.142.5, Auggie 0.31.0, Antigravity CLI 1.0.14, Claude Code 2.1.198; snapshot re-baselined.
-- **Test count**: 1203 → 1208.
+- **Current model IDs**: `scripts/_common.py` now resolves `opus → claude-opus-4-8` and `sonnet → claude-sonnet-5` (the single source of truth generators emit); stale `claude-opus-4-7` samples across skills refreshed.
+- **Effort-aware routing**: `model-routing-patterns` gains an Effort section — tuning `output_config.effort` is the cheaper lever before swapping models, and it does not invalidate the prompt cache the way a mid-session model swap does.
+- **Fable 5 tier documented**: added to the routing table with pricing and a "not the default best model" caveat — for "strongest model", the target stays `claude-opus-4-8`.
+- **Cheaper retrieval agents**: `explorer-agent` and `fact-checker` (pure read/search) moved to `model: haiku`, closing the gap where the skill recommended Haiku workers but no agent used the tier.
 
 See [CHANGELOG.md](CHANGELOG.md) for full history.
 
@@ -74,10 +73,25 @@ ai-toolkit update --local                         # auto-detects editors
 
 ```bash
 ai-toolkit plugin list                            # show available packs
-ai-toolkit plugin install --editor all --all      # install all for Claude + Codex
+ai-toolkit plugin install --editor all --all      # install all for Claude Code + Codex
 ai-toolkit plugin status --editor all             # show what's installed
 ```
 
+### Claude Chat / Desktop / Cowork
+
+The Claude app does not read Claude Code's `~/.claude/rules/` or `CLAUDE.md`.
+Export and upload the app-native plugin instead:
+
+```bash
+ai-toolkit claude-app export --verify
+# Claude > Customize > Plugins > + > Upload plugin
+# Paste the generated *-global-instructions.md into:
+# Settings > Cowork > Global instructions
+```
+
+Re-export and re-upload after toolkit or registered-rule updates. Skills work
+in Chat and Cowork; hooks and sub-agents are active only in Cowork.
+
 ### Install Profiles
 
 ```bash
@@ -102,8 +116,9 @@ See [CLI Reference](kb/reference/cli-reference.md) for all commands and options.
 | Platform | Config Files | Hooks | Scope |
 |----------|-------------|:-----:|-------|
 | Claude Code | `~/.claude/agents`, `~/.claude/skills`, `~/.claude/rules/*.md`, `~/.claude/settings.json` | ✅ | global |
+| Claude Chat / Cowork | uploaded plugin ZIP + UI global/folder instructions | Cowork only | account/app |
 | Cursor | `.cursor/rules/*.mdc` + `.cursor/mcp.json` + `.cursor/skills/*` | ✅ | project (`~/.cursor/mcp.json` for MCP only) |
-| Windsurf (Devin Desktop) | `~/.codeium/.../global_rules.md` + `~/.codeium/windsurf/skills/*` + `.devin/rules/*.md` + `.windsurf/rules/*.md` (legacy) | ✅ | global + project |
+| Windsurf (Devin Desktop) | `~/.config/devin/AGENTS.md` + `.devin/rules/*.md` + `.devin/hooks.v1.json` + `.windsurf/skills/*` | ✅ | global + project |
 | Gemini CLI | `~/.gemini/GEMINI.md` | ✅ | global |
 | GitHub Copilot | `.github/copilot-instructions.md` + `.github/instructions/*` + `.github/prompts/*` + `AGENTS.md` | — | project |
 | Cline | `~/Documents/Cline/Rules/*.md` + `~/.cline/skills/*` + `.clinerules/*.md` | — | global + project |
@@ -114,7 +129,7 @@ See [CLI Reference](kb/reference/cli-reference.md) for all commands and options.
 | Codex CLI | `AGENTS.md` (coding rules inlined) + `.agents/skills/*` + `.codex/hooks.json` | ✅ | project + global plugin |
 | opencode | `AGENTS.md` + `.opencode/{agents,commands,plugins}/*` + `opencode.json` | ✅ | project + global (`~/.config/opencode/`) |
 
-> Claude Code is always installed (primary platform). Other editors on demand with `--editors`. Every platform receives the agent/skill catalog, guidelines, and registered custom rules as text. The **Hooks** column marks platforms that also get lifecycle hook enforcement — the machine-enforced constitution (guard-destructive, quality gates, search-first discipline). Platforms marked — receive those rules as guidance only, without blocking hooks.
+> Claude Code is always installed (primary platform). Other editors are selected with `--editors`; the Claude app uses the separate `claude-app export` flow because its customization store is UI/plugin-managed. The **Hooks** column marks platforms with lifecycle enforcement. Platforms marked — receive guidance without blocking hooks.
 
 ---
 
@@ -126,7 +141,7 @@ See [CLI Reference](kb/reference/cli-reference.md) for all commands and options.
 | `skills/` (hybrid) | 30 | Slash commands with agent knowledge base |
 | `skills/` (knowledge) | 46 | Domain knowledge auto-loaded by agents (includes 13 `<lang>-rules` skills) |
 | `agents/` | 44 | Specialized agents across 10 categories |
-| `hooks/` | 29 entries / 14 events | Quality gates, path safety, prompt governance, loop guard, session lifecycle |
+| `hooks/` | 28 entries / 14 events + statusLine | Quality gates, path safety, prompt governance, loop guard, session lifecycle |
 | `plugins/` | 11 packs | Opt-in domain bundles (security, research, frontend, enterprise, 6 language packs) |
 | `constitution.md` | 7 articles | Machine-enforced safety rules |
 | `rules/` | auto-synced | Global/project rule files for Claude and other editors |
@@ -142,14 +157,15 @@ ai-toolkit/
 │   ├── agents/          # 44 agent definitions
 │   ├── skills/          # 108 skills (task / hybrid / knowledge)
 │   ├── rules/           # Source rules synced into Claude/editor rule files
-│   ├── hooks/           # Hook scripts (29 entries, 14 lifecycle events)
+│   ├── hooks/           # Hook scripts (28 entries, 14 lifecycle events)
+│   ├── claude-app/      # Generated Chat/Cowork plugin rules, hooks, instructions
 │   ├── plugins/         # 11 experimental plugin packs (opt-in)
 │   ├── output-styles/   # System prompt output style overrides
 │   ├── constitution.md  # 7 immutable safety articles
 │   └── ARCHITECTURE.md  # Full system design
 ├── kb/                  # Reference docs, procedures, plans
 ├── scripts/             # Validation, install, evaluation scripts
-├── tests/               # Bats test suite (1208 tests)
+├── tests/               # Bats test suite (1216 tests)
 └── CHANGELOG.md
 ```
 

package/app/.claude-plugin/plugin.json

@@ -1,19 +1,18 @@
 {
+  "$schema": "https://json.schemastore.org/claude-code-plugin-manifest.json",
   "name": "ai-toolkit",
-  "description": "Professional-grade Claude Code toolkit with persona presets, skill security auditor, expanded lifecycle hooks, experimental opt-in plugin packs, benchmark harvesting, and multi-tool support.",
-  "version": "4.12.0",
+  "displayName": "AI Toolkit",
+  "description": "Professional-grade engineering skills, agents, rules, and lifecycle guardrails for Claude Code, Claude Chat, and Cowork.",
+  "version": "4.14.0",
   "author": {
     "name": "SoftSpark",
     "url": "https://github.com/softspark"
   },
-  "repository": {
-    "type": "git",
-    "url": "https://github.com/softspark/ai-toolkit"
-  },
+  "repository": "https://github.com/softspark/ai-toolkit",
   "license": "MIT",
   "homepage": "https://github.com/softspark/ai-toolkit",
-  "npm": "@softspark/ai-toolkit",
-  "install": "npx @softspark/ai-toolkit install",
+  "skills": "./claude-app/skills/",
+  "hooks": "./claude-app/hooks/hooks.json",
   "keywords": [
     "development",
     "testing",
@@ -31,15 +30,8 @@
     "agents",
     "cursor",
     "codex",
-    "opencode"
-  ],
-  "marketplace": {
-    "category": "developer-tools",
-    "tags": ["productivity", "code-quality", "agents", "safety"],
-    "listing-url": "https://smithery.ai/plugin/ai-toolkit",
-    "pricing": "free"
-  },
-  "claude-code": {
-    "min-version": "1.0.33"
-  }
+    "opencode",
+    "cowork",
+    "claude-app"
+  ]
 }

package/app/ARCHITECTURE.md

@@ -8,7 +8,7 @@ Universal multi-agent system for software development. Works across all reposito
 |-----------|-------|
 | Agents | See agents catalog |
 | Skills | See skills catalog |
-| Hooks | 14 events / 29 entries (SessionStart ×3, Notification ×1, PreToolUse ×5, UserPromptSubmit ×2, PostToolUse ×5, Stop ×4, TaskCompleted ×1, TeammateIdle ×1, SubagentStart ×1, SubagentStop ×1, PreCompact ×2, SessionEnd ×1, InstructionsLoaded ×1, ConfigChange ×1) |
+| Hooks | 14 events / 28 entries (SessionStart ×2, Notification ×1, PreToolUse ×5, UserPromptSubmit ×2, PostToolUse ×5, Stop ×4, TaskCompleted ×1, TeammateIdle ×1, SubagentStart ×1, SubagentStop ×1, PreCompact ×2, SessionEnd ×1, InstructionsLoaded ×1, ConfigChange ×1) plus statusLine |
 
 ---
 
@@ -19,8 +19,8 @@ The toolkit is organised as five stacked layers. Higher layers depend on lower l
 ```
 ┌─────────────────────────────────────────────────────────────┐
 │  Layer 5 — Plugins (distribution)                           │
-│  Bundles for npm / marketplace / ai-toolkit eject           │
-│  Files: app/plugins/, plugin.json, package.json             │
+│  Bundles for npm / Claude app / marketplace / eject         │
+│  Files: app/claude-app/, app/plugins/, plugin.json          │
 ├─────────────────────────────────────────────────────────────┤
 │  Layer 4 — Subagents (delegation)                           │
 │  Specialised personas spawned via Agent / Teams             │
@@ -48,14 +48,14 @@ The toolkit is organised as five stacked layers. Higher layers depend on lower l
 | A reusable workflow / domain knowledge | Layer 2 (Skills) | `/debug`, `/review`, `python-rules` |
 | Behavior the harness must enforce automatically | Layer 3 (Hooks) | block destructive `rm -rf`, run lint on Stop |
 | A specialised role for delegation | Layer 4 (Subagents) | `security-auditor`, `database-architect` |
-| A shippable bundle for other repos | Layer 5 (Plugins) | `app/plugins/python-stack/` |
+| A shippable bundle for other repos or Claude app | Layer 5 (Plugins) | `app/plugins/`, `app/claude-app/` |
 
 ### Cross-cutting rules
 
 - A skill MAY trigger an agent (`agent: <name>` frontmatter) — that's Layer 2 calling Layer 4.
 - A hook MUST NOT call a skill — hooks are deterministic shell, skills are LLM-driven.
 - A plugin MAY bundle skills + agents + hooks, but never the inverse.
-- CLAUDE.md is read by every layer; it is the only layer everyone reads.
+- Claude Code reads CLAUDE.md across layers. Claude Chat/Cowork uses app-global/folder instructions and plugin skills instead; `ai-toolkit claude-app export` (backed by `claude_app.py`) builds the uploadable plugin ZIP + Cowork global-instructions and renders core rules into `ai-toolkit-rules` for that runtime.
 
 ---
 
@@ -353,7 +353,7 @@ Lead Session (You)
 ### Language Rules (68 files, 13 languages)
 `app/rules/` contains per-language coding rules. Supported languages: TypeScript, Python, Go, Rust, Java, Kotlin, Swift, Dart, C#, PHP, C++, Ruby, and common (shared). Auto-detected from project files via `--auto-detect` or selected with `--modules rules-<lang>`.
 
-Language rules are propagated to **all configured editors** — not just Claude. `dir_rules_shared.build_language_rules()` reads `app/rules/<lang>/*.md`, strips frontmatter, and returns combined content per language. Each directory-based generator (Cursor, Windsurf, Cline, Roo, Augment, Antigravity, Codex) emits `ai-toolkit-lang-<lang>` files in its native format. Registered custom rules (`~/.softspark/ai-toolkit/rules/`) are similarly propagated as `ai-toolkit-custom-<name>` files via `build_registered_rules()`.
+Language rules are propagated to **all configured editors** — not just Claude Code. `dir_rules_shared.build_language_rules()` reads `app/rules/<lang>/*.md`, strips frontmatter, and returns combined content per language. Each directory-based generator (Cursor, Devin/Windsurf, Cline, Roo, Augment, Antigravity, Codex) emits `ai-toolkit-lang-<lang>` files in its native format. Claude Chat/Cowork receives the existing language skills plus an app-only `ai-toolkit-rules` skill through `claude_app.py`. Registered custom rules are embedded in the local app bundle and propagated to directory-based editors.
 
 ### Codex Integration
 Codex receives `AGENTS.md`, `.agents/rules/*.md`, optional `.agents/skills/*`, and `.codex/hooks.json`. `generate_codex_hooks.py` emits only Codex-supported lifecycle events and prefixes commands with `AI_TOOLKIT_HOOK_QUIET=1`, so informational hook output is not shown at session start or prompt submit while side effects and blocking Stop decisions still run. The `UserPromptSubmit` governance hook also sets `AI_TOOLKIT_HOOK_FORMAT=json` so proactive `additionalContext` reaches the model quietly. Claude's bundled `UserPromptSubmit` hook uses the same quiet JSON context path.

package/app/agents/explorer-agent.md

@@ -2,7 +2,7 @@
 name: explorer-agent
 description: "Codebase exploration and discovery agent. Use for mapping project structure, finding dependencies, understanding architecture, and research. Does NOT write code - only reads and analyzes."
 tools: Read, Grep, Glob
-model: sonnet
+model: haiku
 color: cyan
 skills: clean-code
 ---

package/app/agents/fact-checker.md

@@ -1,7 +1,7 @@
 ---
 name: fact-checker
 description: "Claim verification expert. Use for verifying facts, source validation, RAG result accuracy checking. Triggers: fact check, verify, accuracy, claim, source validation."
-model: sonnet
+model: haiku
 color: cyan
 tools: Read
 skills: clean-code

package/app/claude-app/global-instructions.md

@@ -0,0 +1,10 @@
+# AI Toolkit — Claude app global instructions
+
+For every technical, coding, debugging, review, architecture, DevOps, data,
+or file-editing task, load and follow the installed `ai-toolkit-rules` skill
+before acting. Load the relevant language-rule and workflow skills as needed.
+Use ai-toolkit sub-agents for specialist work in Cowork. Before claiming a
+task is complete, run fresh verification and report the evidence.
+
+In regular Chat, hooks and sub-agents are unavailable; skills remain the
+authoritative customization surface. In Cowork, also honor the plugin hooks.

package/app/claude-app/hooks/hooks.json

@@ -0,0 +1,284 @@
+{
+  "hooks": {
+    "SessionStart": [
+      {
+        "matcher": "startup|compact",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "AI_TOOLKIT_DIR=\"${CLAUDE_PLUGIN_ROOT}\" \"${CLAUDE_PLUGIN_ROOT}/hooks/session-start.sh\""
+          }
+        ]
+      },
+      {
+        "matcher": "startup|compact",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "AI_TOOLKIT_DIR=\"${CLAUDE_PLUGIN_ROOT}\" \"${CLAUDE_PLUGIN_ROOT}/hooks/mcp-health.sh\""
+          }
+        ]
+      }
+    ],
+    "Notification": [
+      {
+        "matcher": "",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "AI_TOOLKIT_DIR=\"${CLAUDE_PLUGIN_ROOT}\" \"${CLAUDE_PLUGIN_ROOT}/hooks/notify-waiting.sh\""
+          }
+        ]
+      }
+    ],
+    "PreToolUse": [
+      {
+        "matcher": "Bash",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "AI_TOOLKIT_DIR=\"${CLAUDE_PLUGIN_ROOT}\" \"${CLAUDE_PLUGIN_ROOT}/hooks/guard-destructive.sh\""
+          }
+        ]
+      },
+      {
+        "matcher": "Bash|Read|Edit|Write|MultiEdit|Glob|Grep|NotebookEdit|mcp__filesystem__.*",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "AI_TOOLKIT_DIR=\"${CLAUDE_PLUGIN_ROOT}\" \"${CLAUDE_PLUGIN_ROOT}/hooks/guard-path.sh\""
+          }
+        ]
+      },
+      {
+        "matcher": "Edit|Write|MultiEdit",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "AI_TOOLKIT_DIR=\"${CLAUDE_PLUGIN_ROOT}\" \"${CLAUDE_PLUGIN_ROOT}/hooks/guard-config.sh\""
+          }
+        ]
+      },
+      {
+        "matcher": "Bash",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "AI_TOOLKIT_DIR=\"${CLAUDE_PLUGIN_ROOT}\" \"${CLAUDE_PLUGIN_ROOT}/hooks/commit-quality.sh\""
+          }
+        ]
+      },
+      {
+        "matcher": "Bash",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "AI_TOOLKIT_DIR=\"${CLAUDE_PLUGIN_ROOT}\" \"${CLAUDE_PLUGIN_ROOT}/hooks/revert-guard.sh\""
+          }
+        ]
+      }
+    ],
+    "UserPromptSubmit": [
+      {
+        "matcher": "",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "AI_TOOLKIT_DIR=\"${CLAUDE_PLUGIN_ROOT}\" AI_TOOLKIT_HOOK_QUIET=1 AI_TOOLKIT_HOOK_FORMAT=json \"${CLAUDE_PLUGIN_ROOT}/hooks/user-prompt-submit.sh\""
+          }
+        ]
+      },
+      {
+        "matcher": "",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "AI_TOOLKIT_DIR=\"${CLAUDE_PLUGIN_ROOT}\" \"${CLAUDE_PLUGIN_ROOT}/hooks/track-usage.sh\""
+          }
+        ]
+      }
+    ],
+    "PostToolUse": [
+      {
+        "matcher": "Edit|MultiEdit|Write",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "AI_TOOLKIT_DIR=\"${CLAUDE_PLUGIN_ROOT}\" \"${CLAUDE_PLUGIN_ROOT}/hooks/post-tool-use.sh\""
+          }
+        ]
+      },
+      {
+        "matcher": "Bash|Edit|Write|MultiEdit",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "AI_TOOLKIT_DIR=\"${CLAUDE_PLUGIN_ROOT}\" \"${CLAUDE_PLUGIN_ROOT}/hooks/governance-capture.sh\""
+          }
+        ]
+      },
+      {
+        "matcher": "Bash|Edit|MultiEdit|Write",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "AI_TOOLKIT_DIR=\"${CLAUDE_PLUGIN_ROOT}\" \"${CLAUDE_PLUGIN_ROOT}/hooks/loop-guard.sh\""
+          }
+        ]
+      },
+      {
+        "matcher": "Edit|MultiEdit|Write",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "AI_TOOLKIT_DIR=\"${CLAUDE_PLUGIN_ROOT}\" \"${CLAUDE_PLUGIN_ROOT}/hooks/test-cohesion.sh\""
+          }
+        ]
+      },
+      {
+        "matcher": "mcp__.*__(smart_query|hybrid_search_kb|crag_search|multi_hop_search|verify_answer)|WebSearch|WebFetch",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "AI_TOOLKIT_DIR=\"${CLAUDE_PLUGIN_ROOT}\" \"${CLAUDE_PLUGIN_ROOT}/hooks/search-tracker.sh\""
+          }
+        ]
+      }
+    ],
+    "Stop": [
+      {
+        "matcher": "",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "AI_TOOLKIT_DIR=\"${CLAUDE_PLUGIN_ROOT}\" \"${CLAUDE_PLUGIN_ROOT}/hooks/quality-check.sh\""
+          }
+        ]
+      },
+      {
+        "matcher": "",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "AI_TOOLKIT_DIR=\"${CLAUDE_PLUGIN_ROOT}\" \"${CLAUDE_PLUGIN_ROOT}/hooks/save-session.sh\""
+          }
+        ]
+      },
+      {
+        "matcher": "",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "AI_TOOLKIT_DIR=\"${CLAUDE_PLUGIN_ROOT}\" \"${CLAUDE_PLUGIN_ROOT}/hooks/quality-gate.sh\""
+          }
+        ]
+      },
+      {
+        "matcher": "",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "AI_TOOLKIT_DIR=\"${CLAUDE_PLUGIN_ROOT}\" \"${CLAUDE_PLUGIN_ROOT}/hooks/stop-search-check.sh\""
+          }
+        ]
+      }
+    ],
+    "TaskCompleted": [
+      {
+        "matcher": "",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "AI_TOOLKIT_DIR=\"${CLAUDE_PLUGIN_ROOT}\" \"${CLAUDE_PLUGIN_ROOT}/hooks/quality-gate.sh\""
+          }
+        ]
+      }
+    ],
+    "TeammateIdle": [
+      {
+        "matcher": "",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "echo 'Teammate going idle. Verify: (1) All assigned files modified, (2) Tests written for changes, (3) Documentation updated if needed.'"
+          }
+        ]
+      }
+    ],
+    "SubagentStart": [
+      {
+        "matcher": "",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "AI_TOOLKIT_DIR=\"${CLAUDE_PLUGIN_ROOT}\" \"${CLAUDE_PLUGIN_ROOT}/hooks/subagent-start.sh\""
+          }
+        ]
+      }
+    ],
+    "SubagentStop": [
+      {
+        "matcher": "",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "AI_TOOLKIT_DIR=\"${CLAUDE_PLUGIN_ROOT}\" \"${CLAUDE_PLUGIN_ROOT}/hooks/subagent-stop.sh\""
+          }
+        ]
+      }
+    ],
+    "PreCompact": [
+      {
+        "matcher": "",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "AI_TOOLKIT_DIR=\"${CLAUDE_PLUGIN_ROOT}\" \"${CLAUDE_PLUGIN_ROOT}/hooks/pre-compact.sh\""
+          }
+        ]
+      },
+      {
+        "matcher": "",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "AI_TOOLKIT_DIR=\"${CLAUDE_PLUGIN_ROOT}\" \"${CLAUDE_PLUGIN_ROOT}/hooks/pre-compact-save.sh\""
+          }
+        ]
+      }
+    ],
+    "SessionEnd": [
+      {
+        "matcher": "",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "AI_TOOLKIT_DIR=\"${CLAUDE_PLUGIN_ROOT}\" \"${CLAUDE_PLUGIN_ROOT}/hooks/session-end.sh\""
+          }
+        ]
+      }
+    ],
+    "InstructionsLoaded": [
+      {
+        "matcher": "",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "AI_TOOLKIT_DIR=\"${CLAUDE_PLUGIN_ROOT}\" \"${CLAUDE_PLUGIN_ROOT}/hooks/instructions-audit.sh\""
+          }
+        ]
+      }
+    ],
+    "ConfigChange": [
+      {
+        "matcher": "user_settings",
+        "hooks": [
+          {
+            "type": "command",
+            "command": "AI_TOOLKIT_DIR=\"${CLAUDE_PLUGIN_ROOT}\" \"${CLAUDE_PLUGIN_ROOT}/hooks/config-desync-guard.sh\""
+          }
+        ]
+      }
+    ]
+  }
+}