@softspark/ai-toolkit 4.5.1 → 4.7.0

package/AGENTS.md

@@ -597,6 +597,56 @@ Derived from the immutable safety constitution (6 articles):
 - **Cite Sources**: Always reference `[PATH: ...]` when making decisions based on existing knowledge
 - **Read-Only Exploration**: Discovery agents never write; writing agents never explore blindly
 - **No Secrets in Code**: Never commit credentials, API keys, or sensitive configuration values
+
+## Coding Rules
+
+### Code Style
+
+* Follow language-specific conventions: PEP 8 (Python), StandardJS/Prettier (TypeScript), gofmt (Go), rustfmt (Rust)
+* Use descriptive names: functions as verbs (`calculateTotal`), booleans as questions (`isValid`), constants as UPPER_SNAKE
+* Keep functions short — single responsibility, max ~30 lines
+* Prefer immutability: use `const`/`final`/`let` over mutable variables where possible
+* No magic numbers — extract to named constants
+* Avoid deep nesting (max 3 levels) — use early returns and guard clauses
+* DRY: extract shared logic only when used 3+ times; premature abstraction is worse than duplication
+* YAGNI: do not build features or abstractions for hypothetical future requirements
+
+### Testing
+
+* Every new feature or bug fix must include tests
+* Use Arrange-Act-Assert pattern for unit tests
+* Test behavior, not implementation — tests should survive refactoring
+* Use descriptive test names: `test_<what>_<when>_<expected>`
+* Prefer real dependencies over mocks at integration boundaries
+* Target >70% code coverage for new code
+* Never skip or disable tests without a linked issue explaining why
+* Run the full test suite before marking work as done
+
+### Security
+
+* Never commit secrets, API keys, credentials, or tokens — use environment variables
+* Validate and sanitize all external input (user input, API responses, file uploads)
+* Use parameterized queries — never concatenate SQL strings
+* Escape output to prevent XSS in web contexts
+* Apply principle of least privilege for file permissions and API scopes
+* Keep dependencies updated — audit regularly for known CVEs
+* Use HTTPS for all external communication
+* Log security events without logging sensitive data (passwords, tokens, PII)
+
+### Output Mode
+
+`output-mode: concise`
+
+Default response mode is **concise**. The `brand-voice` skill (when present) auto-loads concise rules; assistants without that skill should still apply the directives below.
+
+* No preamble — skip "I'll now...", "Sure, let me...", "Great question!" Start with the answer.
+* Lead with the result — conclusion or output first; explanation only if asked or non-obvious.
+* Max 3 sentences per closed question — yes/no, single-fact, or "where is X" answers stay under three sentences.
+* Tables and lists over prose — when comparing options, listing steps, or showing values.
+* No trailing summaries — if the diff or output already shows what changed, do not restate it.
+* Drop filler adjectives — no "nice", "great", "powerful", "robust" unless the user asked for evaluation.
+* Cite as `path:line` — instead of paragraphs describing where things live.
+* Escalate to verbose only for: architecture / RFC / ADR / trade-off documents, or when the user asks for detail.
 <!-- TOOLKIT:ai-toolkit END -->
 
 <!-- TOOLKIT:jira-mcp START -->

package/CHANGELOG.md

@@ -7,6 +7,52 @@ Versioning follows [Semantic Versioning](https://semver.org/).
 
 ---
 
+## v4.7.0 - Devin Desktop .devin tree + Antigravity CLI surfaces (2026-06-10)
+
+Minor release. Ecosystem sync per `kb/procedures/ecosystem-sync-sop.md` (window 2026-06-05 → 2026-06-10, doctor run + per-tool docs review with adversarial verification). No skill/agent count change; no new broad-access skills.
+
+### Ecosystem (class B/D/F drift)
+- **Windsurf/Devin Desktop: `.devin/` dual-emit** — `generate_windsurf_rules.py` and `generate_windsurf_skills.py` now write rules, workflows, and the skill pointer to `.devin/{rules,workflows,skills}/` (primary tree since the 2026-06-02 rebrand) alongside legacy `.windsurf/` (still read as fallback). `.devin/rules` added to editor auto-detection. Test count: 1179 → 1186.
+- **Antigravity: CLI skill pointer** — `generate_antigravity.py` dual-emits the skill pointer to `.agent/skills/` (IDE) and `.agents/skills/` (CLI, plural). Editor auto-detection no longer misreads a pointer-only `.agents/skills/` as a Codex install.
+- **Registry coverage** — added verified config surfaces to `scripts/ecosystem_tools.json`: Claude Code project `.claude/settings.json` + `.claude/rules/*.md` + `http` hook handler; Cursor subagent fallback dirs; Devin Local config files; Gemini CLI hook events (`BeforeToolSelection`, `AfterModel`, `Notification`, `PreCompress`) + `~/.agents/skills` alias; Cline CLI/SDK unified `.cline/` layout, `AGENTS.md` rules sources, and plugins; Augment workspace settings + `Notification` event; Codex project-layer `.codex/config.toml`; opencode skill discovery aliases; Antigravity CLI `.agents/{skills,hooks.json,mcp_config.json}`.
+
+### Changed
+- **Cascade deprecation note** — `.windsurf/hooks.json` (Cascade-scoped) marked deprecated: Cascade is available only through 2026-07-01; migration target is the Devin CLI lifecycle-hooks surface (tracked in registry status_note).
+- **Cursor registry fix** — removed phantom `.cursor/rules/*.md` path (plain `.md` files are ignored by Cursor's rules system); fixed the stale claim in `generate_cursor_mdc.py` docstring.
+- **Augment registry fix** — `.augment/guidelines.md` replaced with documented `.augment-guidelines` + `~/.augment/user-guidelines.md`; `our_generators` synced to all 6 augment generators.
+- **Maintenance SOP** — added `claude --safe-mode` (v2.1.169) isolation step to the rule-enforcement troubleshooting flow.
+- Doctor snapshot refreshed (Claude Code 2.1.170, Codex CLI 0.138.0); class-C items recorded as not adopted (`fallbackModel`, Copilot `excludeAgent`); opencode v1.16.0 experimental v2 skill registry tracked as a watch item.
+
+---
+
+## v4.6.0 - Editor ecosystem sync + Codex rules in AGENTS.md (2026-06-05)
+
+Minor release. Syncs the editor registry with current upstream reality (Windsurf→Devin rebrand, Roo Code archive, Gemini CLI sunset, moved docs hosts) and fixes several generators that wrote to paths their editors never read. Headline: Codex now actually receives the universal coding rules.
+
+### Added
+- **Codex CLI: all 10 native hook events** — wired the 5 previously-missing events (`PostToolUse`, `PreCompact`, `PostCompact`, `SubagentStart`, `SubagentStop`) to the shared toolkit hook scripts, mirroring `app/hooks.json`. Was 5 of 10.
+
+### Changed
+- **Codex CLI: coding rules now in `AGENTS.md`** — the universal code-style/testing/security/output-mode rule bodies are inlined into `AGENTS.md`, the only instruction path Codex reads. They were previously emitted to `.agents/rules/`, which Codex never reads (confirmed against `developers.openai.com/codex`). Language rules continue via `.agents/skills/`.
+- **Google Antigravity: plural `.agents/`** — generators now emit `.agents/rules/` and `.agents/workflows/` (the Antigravity 2.0 default); singular `.agent/` is still read by Antigravity as a fallback.
+- **Conditional skill pointers** — Cursor/Cline/Augment skip the `ai-toolkit-skill-catalogue` pointer when real skills are discoverable at `.claude/skills/` (project or `~/.claude/skills/`), emitting it only as the editor-only fallback. Windsurf keeps its pointer unconditionally (its `.claude/skills` scan is gated behind a Devin "Claude Code config reading" setting).
+- **Cursor subagents** — emit `model: inherit`; dropped the undocumented `tools`/`color` fields to match the current Cursor schema.
+- **Augment subagents** — omit the `model` field (Augment uses the CLI default model; `inherit` is not a documented value).
+- **Ecosystem registry sync** — Windsurf → Devin Desktop (docs/changelog redirects); Roo Code repo archived + docs moved to `roocodeinc.github.io`; Claude Code docs host `code.claude.com`; Codex docs `developers.openai.com/codex`; Gemini CLI free/paid tier sunset 2026-06-18 (enterprise continues); Cline config-path corrections; stale Gemini `Stop` marker dropped.
+
+### Fixed
+- **opencode commands were empty** — the prompt now lives in the markdown body. Commands previously put it in a `template:` frontmatter block (JSON-config-only) with an empty body, so every generated opencode command was effectively blank.
+- **Cline global rules** — now written to `~/Documents/Cline/Rules/` (the documented Cline global path) instead of the never-read `~/.cline/rules/`.
+- **Augment commands** — dropped the undocumented `agent:` frontmatter field; fixed a false `generate_augment_hooks.py` docstring claim about per-workspace settings.
+
+### Removed
+- **`scripts/generate_codex_rules.py`** and the `ai-toolkit codex-rules` / `npm run generate:codex-rules` subcommands — Codex never read the `.agents/rules/` output. Coding rules now ship via `AGENTS.md` (universal) and `.agents/skills/` (language). Installs need no migration; if you scripted `ai-toolkit codex-rules`, the rules are now produced by `ai-toolkit codex-md`. Test count: 1181 → 1179 (dead-path tests removed).
+
+### Ecosystem
+- Class B (integrate): Codex 10 hook events. Class E (now-default): Antigravity plural `.agents/`, conditional skill pointers. Class D (deprecation): Windsurf→Devin, Roo Code archive, Gemini CLI tier sunset. Plus registry URL/version/config-path corrections across 9 tools.
+
+---
+
 ## v4.5.1 - ShellCheck fix for loop-guard hook (2026-06-02)
 
 Patch release. Clears two ShellCheck `SC2034` warnings in `loop-guard.sh` that turned the `main` CI ShellCheck job red after v4.5.0. The publish workflow does not run ShellCheck, so v4.5.0 published despite the red CI; this patch makes `main` green again. No runtime behavior change.

package/README.md

@@ -6,17 +6,17 @@
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](LICENSE)
 [![Skills](https://img.shields.io/badge/skills-107-brightgreen)](app/skills/)
 [![Agents](https://img.shields.io/badge/agents-44-blue)](app/agents/)
-[![Tests](https://img.shields.io/badge/tests-1179%20passing-success)](tests/)
+[![Tests](https://img.shields.io/badge/tests-1186%20passing-success)](tests/)
 
-## What's New in v4.5.1
+## What's New in v4.7.0
 
-v4.5.1 is a patch that clears a ShellCheck lint in the loop-guard hook. The v4.5.0 feature set: hardens the security audit against invisible prompt injection, adds a repeated-action loop guard, fixes destructive-guard false positives, and makes the instinct system actually work.
+v4.7.0 follows the Windsurf → Devin Desktop transition to its config surfaces and brings the editor registry up to date with verified upstream docs.
 
-- **Unicode-safety scanner**: the skill/agent audit now flags invisible and smuggled Unicode (tag-block ASCII smuggling, Trojan Source bidi controls) across all shipped prompt text — a prompt-injection vector regex checks never caught. Runs in the `--ci` gate.
-- **Loop guard**: a new advisory `PostToolUse` hook warns when the same action repeats, catching stuck loops that the `/repeat` failure-counter circuit breaker misses.
-- **`git push --force-with-lease` unblocked**: the destructive-command guard no longer false-positives on safe force-pushes or on commit messages that merely mention `DROP TABLE`/`rm -rf`.
-- **Instincts load by default**: hand-authored `.claude/instincts/*.md` now load at session start (previously dormant behind a verbose flag). The skill docs were corrected to drop a never-implemented auto-extractor claim.
-- **Editor honesty + secret scrubbing**: the platform matrix marks which editors get hook enforcement, and memory-pack now redacts secrets before storing session observations.
+- **Devin Desktop `.devin/` tree**: windsurf generators now dual-emit rules, workflows, and the skill pointer to `.devin/` (primary since the 2026-06-02 rebrand) alongside legacy `.windsurf/` (still read as fallback); `.devin/rules` joins editor auto-detection.
+- **Antigravity CLI skill pointer**: dual-emitted to `.agent/skills/` (IDE) and `.agents/skills/` (CLI, plural); auto-detection no longer misreads a pointer-only `.agents/skills/` as a Codex install.
+- **Cascade deprecation**: the Cascade-scoped `.windsurf/hooks.json` surface dies with Cascade on 2026-07-01 — deprecation note added, migration target is Devin CLI lifecycle hooks.
+- **Registry accuracy**: phantom `.cursor/rules/*.md` removed (Cursor ignores plain `.md`), Augment guidelines path corrected to `.augment-guidelines`, plus verified config surfaces added for Claude Code, Gemini CLI, Cline CLI/SDK, Codex, opencode, and Antigravity CLI.
+- **Troubleshooting**: maintenance SOP gains the `claude --safe-mode` (v2.1.169) isolation step.
 
 See [CHANGELOG.md](CHANGELOG.md) for full history.
 
@@ -103,15 +103,15 @@ See [CLI Reference](kb/reference/cli-reference.md) for all commands and options.
 |----------|-------------|:-----:|-------|
 | Claude Code | `~/.claude/` | ✅ | global |
 | Cursor | `.cursor/rules/*.mdc` + `.cursor/mcp.json` + `.cursor/skills/*` | ✅ | project (`~/.cursor/mcp.json` for MCP only) |
-| Windsurf | `~/.codeium/.../global_rules.md` + `~/.codeium/windsurf/skills/*` + `.windsurf/rules/*.md` | ✅ | global + project |
+| Windsurf (Devin Desktop) | `~/.codeium/.../global_rules.md` + `~/.codeium/windsurf/skills/*` + `.devin/rules/*.md` + `.windsurf/rules/*.md` (legacy) | ✅ | global + project |
 | Gemini CLI | `~/.gemini/GEMINI.md` | ✅ | global |
 | GitHub Copilot | `.github/copilot-instructions.md` | — | project |
-| Cline | `~/.cline/rules/*.md` + `~/.cline/skills/*` + `.clinerules/*.md` | — | global + project |
+| Cline | `~/Documents/Cline/Rules/*.md` + `~/.cline/skills/*` + `.clinerules/*.md` | — | global + project |
 | Roo Code | `~/.roo/rules/*.md` + `.roomodes` + `.roo/rules/*.md` | — | global rules + project |
 | Aider | `~/.aider.conf.yml` + `.aider.conf.yml` + `CONVENTIONS.md` | — | global + project |
 | Augment | `~/.augment/rules/*.md` + `.augment/rules/ai-toolkit-*.md` | ✅ | global + project |
-| Google Antigravity | `.agent/rules/*.md` + `.agent/workflows/*.md` | — | project |
-| Codex CLI | `AGENTS.md` + `.agents/rules/*.md` + `.agents/skills/*` + `.codex/hooks.json` | ✅ | project + global plugin |
+| Google Antigravity | `.agents/rules/*.md` + `.agents/workflows/*.md` + skill pointer in `.agent/skills/*` (IDE) and `.agents/skills/*` (CLI) | — | project |
+| 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.
@@ -149,7 +149,7 @@ ai-toolkit/
 │   └── ARCHITECTURE.md  # Full system design
 ├── kb/                  # Reference docs, procedures, plans
 ├── scripts/             # Validation, install, evaluation scripts
-├── tests/               # Bats test suite (1179 tests)
+├── tests/               # Bats test suite (1186 tests)
 └── CHANGELOG.md
 ```
 

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

@@ -1,7 +1,7 @@
 {
   "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": "4.5.1",
+  "version": "4.7.0",
   "author": {
     "name": "SoftSpark",
     "url": "https://github.com/softspark"

package/benchmarks/ecosystem-doctor-snapshot.json

@@ -1,5 +1,5 @@
 {
-  "last_run": "2026-05-30T19:51:17Z",
+  "last_run": "2026-06-09T22:04:06Z",
   "schema_version": 1,
   "tools": {
     "aider": {
@@ -24,7 +24,7 @@
       }
     },
     "augment": {
-      "docs_hash": "46ef1d20d537234f",
+      "docs_hash": "f4c5995e8333a4d5",
       "headings": [
         "Admin",
         "Auggie CLI",
@@ -51,6 +51,7 @@
         "Auggie CLI": true,
         "MCP": true,
         "Next Edit": false,
+        "Notification": true,
         "PostToolUse": false,
         "PreToolUse": false,
         "SKILL.md": false,
@@ -65,7 +66,7 @@
       }
     },
     "claude-code": {
-      "docs_hash": "b2ee6850b58cee40",
+      "docs_hash": "e6f6b4d9d4057cb0",
       "headings": [
         "Core concepts",
         "Documentation Index",
@@ -114,16 +115,17 @@
         "WorktreeRemove": false,
         "hook handler: agent": false,
         "hook handler: command": false,
+        "hook handler: http": false,
         "hook handler: mcp_tool": false,
         "hook handler: prompt": false,
         "output style": true,
         "slash command": true,
         "sub-agent": true
       },
-      "version": "2.1.158 (Claude Code)"
+      "version": "2.1.170 (Claude Code)"
     },
     "cline": {
-      "docs_hash": "9ec25e5132332d2f",
+      "docs_hash": "806fa530c90e227d",
       "headings": [
         "API Reference",
         "Best Practices",
@@ -151,6 +153,7 @@
         "\u200bWhat is Cline?"
       ],
       "markers": {
+        "AGENTS.md": false,
         "Act Mode": true,
         "MCP": true,
         "Plan Mode": false,
@@ -158,43 +161,58 @@
         "conditional rules": false,
         "custom modes": false,
         "hooks": true,
+        "plugins": true,
         "skills": true,
         "subagents": true,
         "workflows": true
       }
     },
     "codex-cli": {
-      "docs_hash": "f7a6a5144e0f61a2",
+      "docs_hash": "1b4a56a8e918d721",
       "headings": [
-        "About",
-        "Contributing",
-        "Contributors",
-        "Docs",
-        "Folders and files",
-        "Footer",
-        "Footer navigation",
-        "Forks",
-        "History",
-        "Installing and running Codex CLI",
-        "Languages",
-        "Latest commit",
-        "License",
-        "Navigation Menu",
-        "Packages 0",
-        "Provide feedback",
+        "API",
+        "API Reference",
+        "Administration",
+        "Advertiser API",
+        "Agents SDK",
+        "Automation",
+        "Build",
+        "Categories",
+        "Codex",
+        "Codex for Open Source",
+        "Community",
+        "Configuration",
+        "Contribute",
+        "Conversion apps",
+        "Core Concepts",
+        "Core concepts",
+        "Deploy",
+        "Docs agent",
+        "Evaluation",
+        "Events",
+        "Explore use cases",
+        "File Upload",
+        "Get started",
+        "Getting Started",
+        "Going live",
+        "Guides",
+        "Learn",
+        "Legacy APIs",
+        "Measurement",
+        "Plan",
+        "Programs",
         "Quickstart",
-        "Releases 804",
-        "Repository files navigation",
+        "Realtime and audio",
+        "Recent",
+        "Releases",
         "Resources",
-        "Saved searches",
-        "Search code, repositories, users, issues, pull requests...",
-        "Security policy",
-        "Stars",
-        "Uh oh!",
-        "Use saved searches to filter your results more quickly",
-        "Using Codex with your ChatGPT plan",
-        "Watchers",
-        "openai/codex"
+        "Run and scale",
+        "Search the Codex docs",
+        "Specialized models",
+        "Suggested",
+        "Tools",
+        "Topics",
+        "Using Codex"
       ],
       "markers": {
         ".agents/skills": false,
@@ -213,14 +231,14 @@
         "hook handler: agent": false,
         "hook handler: command": false,
         "hook handler: prompt": false,
-        "hooks": false,
+        "hooks": true,
         "mcp_servers": false,
         "sandbox": true
       },
-      "version": "codex-cli 0.134.0"
+      "version": "codex-cli 0.138.0"
     },
     "cursor": {
-      "docs_hash": "6aa8a8d7cc046f13",
+      "docs_hash": "b68e4adf6e21f543",
       "headings": [],
       "markers": {
         ".cursor/rules": false,
@@ -236,7 +254,7 @@
       }
     },
     "gemini-cli": {
-      "docs_hash": "5569709cc2e463a8",
+      "docs_hash": "a0e9a213229063c8",
       "headings": [
         "Breadcrumbs",
         "Directory actions",
@@ -257,15 +275,18 @@
       ],
       "markers": {
         "AfterAgent": false,
+        "AfterModel": false,
         "AfterTool": false,
         "BeforeAgent": false,
         "BeforeModel": false,
         "BeforeTool": false,
+        "BeforeToolSelection": false,
         "GEMINI.md": true,
+        "Notification": true,
+        "PreCompress": false,
         "SKILL.md": false,
         "SessionEnd": false,
         "SessionStart": false,
-        "Stop": true,
         "activate_skill": false,
         "custom commands": false,
         "gemini-extension.json": false,
@@ -275,16 +296,17 @@
       }
     },
     "github-copilot": {
-      "docs_hash": "56ac2d5768f930b5",
+      "docs_hash": "ead818d74ead1f2e",
       "headings": [
         "About Copilot auto model selection",
+        "About Copilot automations",
         "About Copilot integrations",
+        "About agent apps",
         "About agent management",
         "About agent skills",
         "About billing for GitHub Copilot in organizations and enterprises",
+        "About cloud and local sandboxes for GitHub Copilot",
         "About custom agents",
-        "About customizing GitHub Copilot responses",
-        "About enterprise accounts for Copilot Business",
         "Articles",
         "Did you find what you needed?",
         "GitHub Copilot",
@@ -307,7 +329,7 @@
       }
     },
     "google-antigravity": {
-      "docs_hash": "f933b31ac7414c29",
+      "docs_hash": "5ea5426c4445e746",
       "headings": [],
       "markers": {
         "AGENTS.md": false,
@@ -317,13 +339,14 @@
         "agent manager": false,
         "agent permissions": false,
         "artifacts": false,
+        "hooks": false,
         "rules": false,
         "skills": false,
         "workflows": false
       }
     },
     "opencode": {
-      "docs_hash": "fe7f7590e4553574",
+      "docs_hash": "f546c8b741790cb8",
       "headings": [
         "Add features",
         "Ask questions",
@@ -383,45 +406,44 @@
       }
     },
     "windsurf": {
-      "docs_hash": "a30041baf564742f",
+      "docs_hash": "91034c064d9f8290",
       "headings": [
-        "Accounts",
         "Advanced",
         "Agent Command Center",
         "App Deploys",
-        "Cascade",
         "Context Awareness",
+        "Devin Desktop",
+        "Devin Local",
         "Documentation Index",
-        "Editor",
         "MCP",
         "Memories",
         "On this page",
         "Recommended Plugins",
-        "Security",
+        "Releases",
         "Terminal",
-        "Troubleshooting",
         "Usage",
-        "Welcome to Windsurf",
+        "Welcome to Devin Desktop",
+        "Windsurf Plugins",
         "Workflows",
         "\u200b1. Select setup flow",
         "\u200b2. Choose editor theme",
         "\u200b3. Sign up / Log in",
         "\u200b4. Let\u2019s Surf!",
         "\u200bCustom App Icons (beta)",
+        "\u200bDevin Desktop Next",
         "\u200bForgot to Import VS Code Configurations?",
         "\u200bIncompatible Extensions",
         "\u200bOnboarding",
         "\u200bSet Up",
         "\u200bThings to Try",
-        "\u200bUninstall Windsurf",
-        "\u200bUpdate Windsurf",
-        "\u200bWindsurf Next"
+        "\u200bUninstall Devin Desktop",
+        "\u200bUpdate Devin Desktop"
       ],
       "markers": {
         "AGENTS.md": true,
         "Cascade": true,
         "MCP": true,
-        "SKILL.md": false,
+        "SKILL.md": true,
         "always_on": false,
         "glob": true,
         "hooks": true,

package/bin/ai-toolkit.js

@@ -98,9 +98,8 @@ const COMMANDS = {
   'conventions-md': 'Generate CONVENTIONS.md for Aider (auto-loaded)',
   'augment-rules': 'Generate .augment/rules/ai-toolkit.md for Augment (legacy)',
   'augment-dir-rules': 'Generate .augment/rules/ai-toolkit-*.md for Augment (recommended)',
-  'antigravity-rules': 'Generate .agent/rules/ and .agent/workflows/ for Google Antigravity',
-  'codex-md': 'Generate AGENTS.md for OpenAI Codex CLI',
-  'codex-rules': 'Generate .agents/rules/ for OpenAI Codex CLI',
+  'antigravity-rules': 'Generate .agents/rules/ and .agents/workflows/ for Google Antigravity',
+  'codex-md': 'Generate AGENTS.md (coding rules inlined) for OpenAI Codex CLI',
   'codex-hooks': 'Generate .codex/hooks.json for OpenAI Codex CLI',
   'opencode-md': 'Generate AGENTS.md for opencode',
   'opencode-agents': 'Generate .opencode/agents/ for opencode (subagents)',
@@ -504,7 +503,6 @@ function handleGenerateAll(_args) {
   run(scriptPath('generate_cline_rules.py'), [CWD]);
   run(scriptPath('generate_roo_rules.py'), [CWD]);
   run(scriptPath('generate_augment_rules.py'), [CWD]);
-  run(scriptPath('generate_codex_rules.py'), [CWD]);
   run(scriptPath('generate_codex_hooks.py'), [CWD]);
   run(scriptPath('generate_opencode_agents.py'), [CWD]);
   run(scriptPath('generate_opencode_commands.py'), [CWD]);
@@ -617,7 +615,6 @@ const SPECIAL_HANDLERS = {
   'roo-dir-rules': (_args) => run(scriptPath('generate_roo_rules.py'), [CWD]),
   'conventions-md': (_args) => { const out = runGenerator('generate_conventions.py'); fs.writeFileSync(path.join(CWD, 'CONVENTIONS.md'), out); console.log('Generated: CONVENTIONS.md'); },
   'augment-dir-rules': (_args) => run(scriptPath('generate_augment_rules.py'), [CWD]),
-  'codex-rules': (_args) => run(scriptPath('generate_codex_rules.py'), [CWD]),
   'codex-hooks': (_args) => run(scriptPath('generate_codex_hooks.py'), [CWD]),
   'opencode-agents': (_args) => run(scriptPath('generate_opencode_agents.py'), [CWD]),
   'opencode-commands': (_args) => run(scriptPath('generate_opencode_commands.py'), [CWD]),

package/kb/procedures/maintenance-sop.md

@@ -3,9 +3,9 @@ title: "SOP: Claude Toolkit Maintenance"
 category: procedures
 service: ai-toolkit
 tags: [sop, maintenance, agents, skills, install]
-version: "3.0.1"
+version: "3.1.0"
 created: "2026-03-23"
-last_updated: "2026-05-25"
+last_updated: "2026-06-10"
 description: "Standard operating procedures for installing, maintaining, and evolving the ai-toolkit."
 ---
 
@@ -211,18 +211,19 @@ Use this when Claude appears to ignore `CLAUDE.md`, `.claude/rules/*.md`, output
    - `https://code.claude.com/docs/en/settings`
    - `https://code.claude.com/docs/en/output-styles`
    - `https://code.claude.com/docs/en/hooks`
-2. **Verify instruction loading.** Run `/memory` in Claude Code and confirm the expected `CLAUDE.md`, `CLAUDE.local.md`, and `.claude/rules/*.md` files are listed. Remember that Claude Code reads `CLAUDE.md`, not `AGENTS.md`, unless `CLAUDE.md` imports it.
-3. **Verify the active output style.** Check `.claude/settings.local.json` or `/config`. Output style changes apply after `/clear` or a new session.
-4. **Inspect installed hooks.** Ensure `~/.claude/settings.json` contains the ai-toolkit `UserPromptSubmit` and `Stop` entries. The governance hook must run with `AI_TOOLKIT_HOOK_QUIET=1 AI_TOOLKIT_HOOK_FORMAT=json` so it injects `additionalContext` without noisy transcript output.
-5. **Reproduce the hook path directly.**
+2. **Isolate with safe mode (v2.1.169+).** Start a session with `claude --safe-mode` (or `CLAUDE_CODE_SAFE_MODE=1`) — it disables hooks, skills, agents, and CLAUDE.md. If the problem persists in safe mode, the cause is upstream, not toolkit customization; if it disappears, bisect toolkit components with `ai-toolkit update --only <component>`.
+3. **Verify instruction loading.** Run `/memory` in Claude Code and confirm the expected `CLAUDE.md`, `CLAUDE.local.md`, and `.claude/rules/*.md` files are listed. Remember that Claude Code reads `CLAUDE.md`, not `AGENTS.md`, unless `CLAUDE.md` imports it.
+4. **Verify the active output style.** Check `.claude/settings.local.json` or `/config`. Output style changes apply after `/clear` or a new session.
+5. **Inspect installed hooks.** Ensure `~/.claude/settings.json` contains the ai-toolkit `UserPromptSubmit` and `Stop` entries. The governance hook must run with `AI_TOOLKIT_HOOK_QUIET=1 AI_TOOLKIT_HOOK_FORMAT=json` so it injects `additionalContext` without noisy transcript output.
+6. **Reproduce the hook path directly.**
    ```bash
    printf '{"session_id":"debug","prompt":"debug this technical rule issue"}' \
      | AI_TOOLKIT_SEARCH_FIRST=strict AI_TOOLKIT_HOOK_QUIET=1 AI_TOOLKIT_HOOK_FORMAT=json \
        ~/.softspark/ai-toolkit/hooks/user-prompt-submit.sh
    ```
    The output must be valid JSON with `hookSpecificOutput.additionalContext`.
-6. **Check corrective enforcement.** If the assistant still skips required research, `stop-search-check.sh` should block Stop with the search-first message. If it does not, inspect `~/.softspark/ai-toolkit/state/search-required-*.flag` and the Codex/Claude transcript logs.
-7. **Repair drift.** Run:
+7. **Check corrective enforcement.** If the assistant still skips required research, `stop-search-check.sh` should block Stop with the search-first message. If it does not, inspect `~/.softspark/ai-toolkit/state/search-required-*.flag` and the Codex/Claude transcript logs.
+8. **Repair drift.** Run:
    ```bash
    ai-toolkit update --only hooks
    python3 scripts/ecosystem_doctor.py --tool claude-code --format text