@softspark/ai-toolkit 2.4.0 → 2.5.0

package/kb/reference/comparison.md

@@ -12,7 +12,7 @@ description: "Feature comparison of ai-toolkit vs other Claude Code toolkits and
 
 | Feature | ai-toolkit | everything-claude-code | wshobson/agents | ruflo |
 |---------|---------------|----------------------|-----------------|-------|
-| Skills | 92 | 100+ | 146 | 20+ |
+| Skills | 93 | 100+ | 146 | 20+ |
 | Agents | 44 | 30+ | 112 | 20+ |
 | Machine-enforced constitution | **Yes** | No (docs only) | No | No |
 | Skill-scoped lifecycle hooks | **Yes** | No | No | No |

package/kb/reference/extension-api.md

@@ -103,6 +103,8 @@ npx @softspark/ai-toolkit inject-hook https://example.com/hooks.json my-tool-hoo
 
 **Safety:** Entries tagged `"_source": "ai-toolkit"` are never modified or removed by this command. External tools cannot affect the toolkit's own hooks. Only HTTPS URLs are accepted.
 
+**Codex propagation:** Codex-compatible events (`SessionStart`, `PreToolUse`, `PostToolUse`, `UserPromptSubmit`, `Stop`) are automatically propagated to `~/.codex/hooks.json`. Non-Codex events are silently skipped. No extra flags needed.
+
 ## remove-hook
 
 Strips all hook entries from `~/.claude/settings.json` that carry a given `_source` tag. If the hook was URL-sourced, also unregisters the URL from `sources.json` and removes the cached file.

package/kb/reference/skills-catalog.md

@@ -23,7 +23,7 @@ All functionality is unified under skills. Task and hybrid skills are user-invoc
 | **2 — Multi-agent workflow** | `/workflow <type>` | Cross-cutting task with known pattern |
 | **3 — Custom parallelism** | `/orchestrate`, `/swarm` | No predefined workflow matches |
 
-## Task Skills (29)
+## Task Skills (30)
 
 Task skills execute a specific action. Invoked via slash commands. `disable-model-invocation: true`.
 
@@ -58,6 +58,8 @@ Task skills execute a specific action. Invoked via slash commands. `disable-mode
 | **prd-to-issues** | `/prd-to-issues` | medium | Break PRD into GitHub issues with vertical slices and HITL/AFK tagging |
 | **skill-audit** | `/skill-audit` | medium | Scan skills and agents for security risks: dangerous patterns, secrets, excessive permissions |
 | **hipaa-validate** | `/hipaa-validate` | medium | Scan codebase for HIPAA compliance issues: PHI exposure, missing audit logging, unencrypted transmission/storage, access control gaps, temp file exposure, and missing BAA references |
+| **a11y-validate** | `/a11y-validate` | medium | Scan codebase for accessibility violations: WCAG 2.1 Level AA, EN 301 549, European Accessibility Act (EAA / EU 2019/882). Covers semantics, keyboard, focus, color contrast, forms, media, ARIA, motion, mobile (React Native + Flutter), and EAA accessibility-statement documentation. |
+| **seo-validate** | `/seo-validate` | medium | Scan codebase for SEO issues: W3C semantics, meta/OG tags, Schema.org, hreflang, Core Web Vitals (LCP/INP/CLS), resource hints, GEO, SPA/SSG/CSR crawlability, technical SEO, accessibility-for-SEO. Framework-aware (Next/Nuxt/Astro/Gatsby/SvelteKit/Remix/Angular/Vue/static HTML). |
 
 ## Hybrid Skills (31)
 

package/llms-full.txt

@@ -56,6 +56,7 @@
 
 ## Skills
 
+- **a11y-validate**: Validate code against accessibility standards: WCAG 2.1 Level AA, EN 301 549, and the European Accessibility Act (EAA / Directive EU 2019/882). Covers semantics, text alternatives, keyboard/focus, color/contrast, forms, media, ARIA, motion, mobile, and EAA documentation requirements. Framework-aware (React/Next/Nuxt/Astro/Gatsby/SvelteKit/Remix/Angular/Vue/React Native/Flutter/static HTML).
 - **agent-creator**: Creates new specialized agents with frontmatter, tool selection, and delegation guidance
 - **analyze**: Analyze code quality, complexity, and patterns
 - **api-patterns**: Loaded when user asks about REST API design or GraphQL patterns
@@ -133,6 +134,7 @@
 - **rust-patterns**: Loaded when user asks about Rust development patterns
 - **search**: Search the knowledge base with semantic and hybrid modes
 - **security-patterns**: Loaded when user asks about security, OWASP, or auth patterns
+- **seo-validate**: Validate code against SEO best practices: W3C semantics, meta/OG tags, Schema.org, hreflang, Core Web Vitals (LCP/INP/CLS), resource hints, GEO, SPA/SSG/CSR crawlability, technical SEO, accessibility-for-SEO. Supports React/Next/Nuxt/Astro/Gatsby/SvelteKit/Remix/Angular/Vue/static HTML.
 - **skill-audit**: Scan skills and agents for security risks: dangerous patterns, secrets, excessive permissions
 - **skill-creator**: Create new skills from templates with guided workflow
 - **subagent-development**: Execute implementation plans using fresh subagents per task with two-stage review: spec compliance first, then code quality. Use when executing plans with independent tasks.
@@ -4456,7 +4458,7 @@ service: ai-toolkit
 tags: [architecture, overview, design, structure]
 version: "1.4.4"
 created: "2026-03-23"
-last_updated: "2026-04-13"
+last_updated: "2026-04-15"
 description: "Architecture of ai-toolkit: directory layout, global install model, editor-aware MCP install, Codex translation layer, skill tiers, and integration with projects."
 ---
 
@@ -4636,7 +4638,7 @@ Three tiers determine how to approach a task:
 
 | Type | Field | Invocation | Count |
 |------|-------|-----------|-------|
-| Task | `disable-model-invocation: true` | User via `/skill` only | 29 |
+| Task | `disable-model-invocation: true` | User via `/skill` only | 30 |
 | Hybrid | (neither) | User via `/skill` + agent knowledge | 31 |
 | Knowledge | `user-invocable: false` | Claude auto-loads | 32 |
 
@@ -4786,7 +4788,7 @@ Severity levels: HIGH (blocks deployment), WARN (should fix), INFO (best practic
 ## Extension Points
 
 ### MCP Templates
-`app/plugins/mcp-templates/` contains 26 ready-to-use MCP server config templates. Opt-in via `ai-toolkit install --modules mcp-templates` or activated automatically with `--profile strict|full`.
+`app/mcp-templates/` contains 26 ready-to-use MCP server config templates. Opt-in via `ai-toolkit install --modules mcp-templates` or activated automatically with `--profile strict|full`.
 
 ### Language Rules
 `app/rules/` provides language-specific rule files covering 13 languages (TypeScript, Python, Go, Rust, Java, Kotlin, Swift, Dart, C#, PHP, C++, Ruby, common). Auto-detected from project files via `--auto-detect` or selectable with `--modules rules-<lang>`. See README.md for current count.
@@ -5137,7 +5139,7 @@ category: reference
 service: ai-toolkit
 tags: [cli, commands, reference, install, update, plugin, mcp]
 created: "2026-04-13"
-last_updated: "2026-04-13"
+last_updated: "2026-04-15"
 description: "Complete CLI reference for all ai-toolkit commands, options, and flags."
 ---
 
@@ -5380,6 +5382,10 @@ This means Claude-only events such as `TaskCompleted`, `TeammateIdle`,
 `SubagentStart`, `SubagentStop`, `PreCompact`, `SessionEnd`, and
 `Notification` are not available in `.codex/hooks.json`.
 
+`inject-hook` automatically propagates Codex-compatible events to
+`~/.codex/hooks.json` (global layer). Non-Codex events are silently skipped.
+`remove-hook` cleans both Claude and Codex targets.
+
 ## Behavioral Limits
 
 Codex wrappers preserve workflow intent, but not every Claude runtime behavior
@@ -5426,7 +5432,7 @@ description: "Feature comparison of ai-toolkit vs other Claude Code toolkits and
 
 | Feature | ai-toolkit | everything-claude-code | wshobson/agents | ruflo |
 |---------|---------------|----------------------|-----------------|-------|
-| Skills | 92 | 100+ | 146 | 20+ |
+| Skills | 93 | 100+ | 146 | 20+ |
 | Agents | 44 | 30+ | 112 | 20+ |
 | Machine-enforced constitution | **Yes** | No (docs only) | No | No |
 | Skill-scoped lifecycle hooks | **Yes** | No | No | No |
@@ -6634,6 +6640,8 @@ npx @softspark/ai-toolkit inject-hook https://example.com/hooks.json my-tool-hoo
 
 **Safety:** Entries tagged `"_source": "ai-toolkit"` are never modified or removed by this command. External tools cannot affect the toolkit's own hooks. Only HTTPS URLs are accepted.
 
+**Codex propagation:** Codex-compatible events (`SessionStart`, `PreToolUse`, `PostToolUse`, `UserPromptSubmit`, `Stop`) are automatically propagated to `~/.codex/hooks.json`. Non-Codex events are silently skipped. No extra flags needed.
+
 ## remove-hook
 
 Strips all hook entries from `~/.claude/settings.json` that carry a given `_source` tag. If the hook was URL-sourced, also unregisters the URL from `sources.json` and removes the cached file.
@@ -8547,7 +8555,7 @@ All functionality is unified under skills. Task and hybrid skills are user-invoc
 | **2 — Multi-agent workflow** | `/workflow <type>` | Cross-cutting task with known pattern |
 | **3 — Custom parallelism** | `/orchestrate`, `/swarm` | No predefined workflow matches |
 
-## Task Skills (29)
+## Task Skills (30)
 
 Task skills execute a specific action. Invoked via slash commands. `disable-model-invocation: true`.
 
@@ -8582,6 +8590,8 @@ Task skills execute a specific action. Invoked via slash commands. `disable-mode
 | **prd-to-issues** | `/prd-to-issues` | medium | Break PRD into GitHub issues with vertical slices and HITL/AFK tagging |
 | **skill-audit** | `/skill-audit` | medium | Scan skills and agents for security risks: dangerous patterns, secrets, excessive permissions |
 | **hipaa-validate** | `/hipaa-validate` | medium | Scan codebase for HIPAA compliance issues: PHI exposure, missing audit logging, unencrypted transmission/storage, access control gaps, temp file exposure, and missing BAA references |
+| **a11y-validate** | `/a11y-validate` | medium | Scan codebase for accessibility violations: WCAG 2.1 Level AA, EN 301 549, European Accessibility Act (EAA / EU 2019/882). Covers semantics, keyboard, focus, color contrast, forms, media, ARIA, motion, mobile (React Native + Flutter), and EAA accessibility-statement documentation. |
+| **seo-validate** | `/seo-validate` | medium | Scan codebase for SEO issues: W3C semantics, meta/OG tags, Schema.org, hreflang, Core Web Vitals (LCP/INP/CLS), resource hints, GEO, SPA/SSG/CSR crawlability, technical SEO, accessibility-for-SEO. Framework-aware (Next/Nuxt/Astro/Gatsby/SvelteKit/Remix/Angular/Vue/static HTML). |
 
 ## Hybrid Skills (31)
 

package/manifest.json

@@ -1,5 +1,5 @@
 {
-  "version": "2.4.0",
+  "version": "2.5.0",
   "components": {
     "agents": {
       "description": "44 specialized agents (orchestrator, backend, frontend, security, devops, etc.)",
@@ -9,7 +9,7 @@
       "tags": ["core", "agents"]
     },
     "skills": {
-      "description": "92 skills (29 task + 31 hybrid + 32 knowledge)",
+      "description": "94 skills (31 task + 31 hybrid + 32 knowledge)",
       "path": "app/skills",
       "target": ".claude/skills",
       "type": "symlink",
@@ -90,7 +90,7 @@
       "default": true
     },
     "skills": {
-      "description": "92 skills (task, hybrid, knowledge)",
+      "description": "94 skills (task, hybrid, knowledge)",
       "default": true
     },
     "rules-common": {

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "@softspark/ai-toolkit",
-  "version": "2.4.0",
-  "description": "Professional-grade AI coding toolkit: 92 skills, 44 agents, multi-platform support (Claude, Cursor, Windsurf, Copilot, Gemini, Cline, Roo Code, Aider, Augment, Google Antigravity, Codex CLI), machine-enforced safety constitution, persona presets, skill security auditor, expanded lifecycle hooks, 11 plugin packs, and benchmark tooling.",
+  "version": "2.5.0",
+  "description": "Professional-grade AI coding toolkit: 94 skills, 44 agents, multi-platform support (Claude, Cursor, Windsurf, Copilot, Gemini, Cline, Roo Code, Aider, Augment, Google Antigravity, Codex CLI), machine-enforced safety constitution, persona presets, skill security auditor, expanded lifecycle hooks, 11 plugin packs, and benchmark tooling.",
   "keywords": [
     "claude",
     "claude-code",

package/scripts/config_cli.py

@@ -96,17 +96,11 @@ def cmd_validate(args: list[str]) -> int:
             print(f"  ✗ Merge failed: {e}")
             return 1
 
-    checks = [
-        ("schema valid", True),
-        ("no forbidden overrides", True),
-        ("constitution articles intact", True),
-    ]
-
+    # Summary — these passed because validation above would have returned 1
+    print(f"  ✓ schema valid")
+    print(f"  ✓ no forbidden overrides")
     if extends:
-        checks.append(("extends resolved", True))
-
-    for label, ok in checks:
-        print(f"  {'✓' if ok else '✗'} {label}")
+        print(f"  ✓ extends resolved")
 
     print("\n  Config valid ✓")
     return 0

package/scripts/config_resolver.py

@@ -299,14 +299,28 @@ def _find_cached_npm(cache_dir: Path) -> Path | None:
 
 
 def _extract_tarball(tarball: Path, dest: Path) -> None:
-    """Extract npm tarball (which has a package/ prefix) to dest."""
+    """Extract npm tarball (which has a package/ prefix) to dest.
+
+    Validates that extracted paths stay within dest to prevent path traversal.
+    Rejects symlinks and absolute paths.
+    """
+    dest_resolved = dest.resolve()
     with tarfile.open(tarball, "r:gz") as tf:
         for member in tf.getmembers():
             # npm tarballs have a "package/" prefix
-            if member.name.startswith("package/"):
-                member.name = member.name[len("package/"):]
-                if member.name:  # skip empty (the "package/" dir itself)
-                    tf.extract(member, dest)
+            if not member.name.startswith("package/"):
+                continue
+            member.name = member.name[len("package/"):]
+            if not member.name:  # skip empty (the "package/" dir itself)
+                continue
+            # Reject symlinks and absolute paths
+            if member.issym() or member.islnk() or member.name.startswith("/"):
+                continue
+            # Path traversal protection
+            target = (dest / member.name).resolve()
+            if not str(target).startswith(str(dest_resolved)):
+                continue
+            tf.extract(member, dest)
 
 
 def _extract_version_from_tarball(filename: str, package_name: str) -> str:
@@ -332,6 +346,10 @@ def _resolve_git(
 ) -> BaseConfig:
     """Resolve from git URL (git+https://...)."""
     url = source.removeprefix("git+")
+    if not url.startswith("https://"):
+        raise ConfigResolverError(
+            f"Only HTTPS git URLs are supported (got: {url.split('://')[0]}://)"
+        )
     cache_key = hashlib.sha256(url.encode()).hexdigest()[:16]
     cache_dir = _cache_root() / "git" / cache_key
 

package/scripts/doctor.py

@@ -10,6 +10,7 @@ Checks:
   6. Planned assets
   7. Benchmark freshness
   8. Stale rules
+  9. URL hook sources
 
 Exit codes:
   0  all checks pass
@@ -27,7 +28,7 @@ from pathlib import Path
 
 sys.path.insert(0, str(Path(__file__).resolve().parent))
 from _common import toolkit_dir
-from paths import HOOKS_DIR as _HOOKS_DIR, RULES_DIR as _RULES_DIR
+from paths import HOOKS_DIR as _HOOKS_DIR, RULES_DIR as _RULES_DIR, EXTERNAL_HOOKS_DIR as _EXTERNAL_HOOKS_DIR
 
 
 # ---------------------------------------------------------------------------
@@ -37,6 +38,7 @@ from paths import HOOKS_DIR as _HOOKS_DIR, RULES_DIR as _RULES_DIR
 CLAUDE_DIR = Path.home() / ".claude"
 HOOKS_DIR = _HOOKS_DIR
 RULES_DIR = _RULES_DIR
+EXTERNAL_HOOKS_DIR = _EXTERNAL_HOOKS_DIR
 BENCHMARK_DASHBOARD = toolkit_dir / "benchmarks" / "ecosystem-dashboard.json"
 
 VALID_EVENTS = frozenset({
@@ -452,19 +454,88 @@ def check_stale_rules(dr: DiagResult, fix_mode: bool) -> None:
     for rule_file in sorted(RULES_DIR.iterdir()):
         # Check for stale symlinks
         if rule_file.is_symlink() and not rule_file.exists():
-            print(f"  WARNING: Stale symlink: {rule_file}")
+            dr.warn(f"Stale symlink: {rule_file}")
             stale += 1
             if fix_mode:
                 rule_file.unlink()
-                print("    FIXED: removed stale symlink")
+                dr.fixed(f"removed stale symlink: {rule_file.name}")
         elif rule_file.is_file() and rule_file.stat().st_size == 0:
-            print(f"  WARNING: Empty rule file: {rule_file}")
+            dr.warn(f"Empty rule file: {rule_file}")
             stale += 1
 
     if stale == 0:
         dr.ok("All rules healthy")
 
 
+# ---------------------------------------------------------------------------
+# Check 9: URL Hook Sources
+# ---------------------------------------------------------------------------
+
+def check_url_hooks(dr: DiagResult, fix_mode: bool) -> None:
+    """Check URL-sourced hook cache integrity."""
+    print()
+    print("## 9. URL Hook Sources")
+
+    sources_file = EXTERNAL_HOOKS_DIR / "sources.json"
+    if not sources_file.is_file():
+        dr.skip("No URL hook sources registered")
+        return
+
+    try:
+        with open(sources_file, encoding="utf-8") as f:
+            data = json.load(f)
+        sources = data.get("hooks", {})
+    except (json.JSONDecodeError, OSError) as exc:
+        dr.fail(f"Corrupt sources.json: {exc}")
+        return
+
+    if not sources:
+        dr.ok("No URL hook sources registered")
+        return
+
+    issues = 0
+    for name, entry in sources.items():
+        cached = EXTERNAL_HOOKS_DIR / f"{name}.json"
+        url = entry.get("url", "")
+
+        if not cached.is_file():
+            dr.warn(f"Missing cached file for '{name}' ({url})")
+            issues += 1
+            if fix_mode:
+                try:
+                    from url_fetch import fetch_url
+                    content = fetch_url(url)
+                    json.loads(content)  # validate
+                    EXTERNAL_HOOKS_DIR.mkdir(parents=True, exist_ok=True)
+                    cached.write_bytes(content)
+                    print(f"    FIXED: re-fetched {name}")
+                except Exception as exc:
+                    print(f"    Could not re-fetch: {exc}")
+            continue
+
+        # Validate cached file is valid JSON with hooks key
+        try:
+            with open(cached, encoding="utf-8") as f:
+                hook_data = json.load(f)
+            if "hooks" not in hook_data:
+                dr.warn(f"Cached file '{name}' missing 'hooks' key")
+                issues += 1
+            else:
+                dr.ok(f"{name} ({url})")
+        except json.JSONDecodeError:
+            dr.warn(f"Corrupt cached file: {cached}")
+            issues += 1
+            if fix_mode:
+                try:
+                    from url_fetch import fetch_url
+                    content = fetch_url(url)
+                    json.loads(content)
+                    cached.write_bytes(content)
+                    print(f"    FIXED: re-fetched {name}")
+                except Exception as exc:
+                    print(f"    Could not re-fetch: {exc}")
+
+
 # ---------------------------------------------------------------------------
 # Main
 # ---------------------------------------------------------------------------
@@ -486,6 +557,7 @@ def main() -> None:
     check_planned_assets(dr)
     check_benchmark_freshness(dr)
     check_stale_rules(dr, fix_mode)
+    check_url_hooks(dr, fix_mode)
 
     # Summary
     print("========================")

package/scripts/hook_sources.py

@@ -80,6 +80,9 @@ def save_sources(hooks_dir: Path | None = None,
 
 def register_url_source(hooks_dir: Path | None, hook_name: str, url: str) -> None:
     """Add or update a URL source entry."""
+    import re
+    if not hook_name or not re.fullmatch(r"[a-zA-Z0-9_-]+", hook_name):
+        raise ValueError(f"Invalid hook name: {hook_name!r}")
     hooks_dir = hooks_dir or EXTERNAL_HOOKS_DIR
     sources = load_sources(hooks_dir)
     sources[hook_name] = {

package/scripts/inject_hook_cli.py

@@ -49,6 +49,9 @@ sys.path.insert(0, str(Path(__file__).resolve().parent))
 # Protected source tag -- this CLI must never touch ai-toolkit's own entries.
 PROTECTED_SOURCE = "ai-toolkit"
 
+# Codex CLI supports only these 5 hook events.
+CODEX_EVENTS = {"SessionStart", "PreToolUse", "PostToolUse", "UserPromptSubmit", "Stop"}
+
 
 # ---------------------------------------------------------------------------
 # JSON helpers (same style as merge-hooks.py)
@@ -179,6 +182,70 @@ def merge_hooks(new_hooks: dict, existing_hooks: dict, source: str) -> dict:
     return merged
 
 
+# ---------------------------------------------------------------------------
+# Codex propagation
+# ---------------------------------------------------------------------------
+
+def _codex_hooks_path(target_dir: str) -> Path:
+    """Return the global Codex hooks.json path."""
+    return Path(target_dir) / ".codex" / "hooks.json"
+
+
+def _filter_codex_events(hooks: dict) -> dict:
+    """Keep only events supported by Codex CLI."""
+    return {event: entries for event, entries in hooks.items()
+            if event in CODEX_EVENTS}
+
+
+def _inject_codex(tagged_hooks: dict, source: str, target_dir: str) -> None:
+    """Propagate hook entries to ~/.codex/hooks.json (Codex global layer).
+
+    Only events in CODEX_EVENTS are propagated. Non-Codex events are silently
+    skipped.
+    """
+    codex_hooks = _filter_codex_events(tagged_hooks)
+    if not codex_hooks:
+        return
+
+    codex_path = _codex_hooks_path(target_dir)
+    codex_path.parent.mkdir(parents=True, exist_ok=True)
+
+    existing: dict = {}
+    if codex_path.is_file():
+        try:
+            data = load_json(str(codex_path))
+            existing = data.get("hooks", {})
+        except (json.JSONDecodeError, OSError):
+            existing = {}
+
+    merged = merge_hooks(codex_hooks, existing, source)
+    save_json(str(codex_path), {"hooks": merged})
+    events = ", ".join(sorted(codex_hooks.keys()))
+    print(f"Propagated to Codex: {codex_path} (events: {events})")
+
+
+def _remove_codex(source_name: str, target_dir: str) -> None:
+    """Remove hook entries from ~/.codex/hooks.json."""
+    codex_path = _codex_hooks_path(target_dir)
+    if not codex_path.is_file():
+        return
+
+    try:
+        data = load_json(str(codex_path))
+    except (json.JSONDecodeError, OSError):
+        return
+
+    existing = data.get("hooks", {})
+    cleaned = strip_source(existing, source_name)
+
+    if cleaned:
+        save_json(str(codex_path), {"hooks": cleaned})
+    else:
+        save_json(str(codex_path), {"hooks": {}})
+
+    print(f"Removed '{source_name}' from Codex: {codex_path}")
+
+
 # ---------------------------------------------------------------------------
 # CLI actions
 # ---------------------------------------------------------------------------
@@ -213,7 +280,7 @@ def _fetch_and_cache(url: str, source: str) -> str:
         sys.exit(2)
 
     if "hooks" not in parsed:
-        print(f"Warning: no 'hooks' key found in URL response", file=sys.stderr)
+        print("Warning: no 'hooks' key found in URL response", file=sys.stderr)
 
     cached_path = EXTERNAL_HOOKS_DIR / f"{source}.json"
     cached_path.write_bytes(data)
@@ -310,6 +377,9 @@ def inject(hooks_file: str, target_dir: str, source_override: str = "") -> None:
     save_json(str(settings_path), settings)
     print(f"Injected hooks from '{source}' into {settings_path}")
 
+    # Propagate Codex-compatible events to ~/.codex/hooks.json
+    _inject_codex(tagged, source, target_dir)
+
 
 def remove(source_name: str, target_dir: str) -> None:
     """Remove all hook entries tagged with *source_name*.
@@ -354,6 +424,9 @@ def remove(source_name: str, target_dir: str) -> None:
     save_json(str(settings_path), settings)
     print(f"Removed hooks with source '{source_name}' from {settings_path}")
 
+    # Remove from Codex global hooks
+    _remove_codex(source_name, target_dir)
+
     # Unregister URL source if present
     try:
         from hook_sources import unregister_source

package/scripts/install.py

@@ -54,7 +54,10 @@ from install_steps.install_state import (
     record_install,
     get_installed_modules,
     get_installed_profile,
+    get_global_editors,
+    record_global_editors,
     print_status,
+    GLOBAL_CAPABLE_EDITORS,
 )
 from install_steps.detect_language import detect_languages
 from install_steps.project_registry import register_project
@@ -708,17 +711,31 @@ def main() -> None:
             profile = cfg["profile"]
 
         lang_modules = [m for m in (resolved_modules or []) if m.startswith("rules-")]
-        editors_arg: str = cfg["editors"]
+        local_editors_arg: str = cfg["editors"]
         install_local_project(rules_dir, dry_run, reset, lang_modules or None,
-                              editors=editors_arg,
+                              editors=local_editors_arg,
                               merged_config=merged_config)
+        installed_eds: list[str] = []  # local install doesn't track global editors
         install_strict_git_hooks(profile, local, dry_run)
     else:
         # Global install
         print_banner(target_dir, rules_dir, profile, only, skip, dry_run,
                      modules=resolved_modules)
         install_claude_code(target_dir, hooks_scripts_dir, rules_dir, only, skip, dry_run)
-        install_ai_tools(target_dir, rules_dir, only, skip, dry_run)
+
+        # Determine global editors: --editors flag > state > default (none)
+        editors_arg: str = cfg["editors"]
+        if editors_arg:
+            if editors_arg == "all":
+                global_eds = list(GLOBAL_CAPABLE_EDITORS)
+            else:
+                global_eds = [e.strip() for e in editors_arg.split(",") if e.strip()]
+        else:
+            # On update: use editors from state; on fresh install: none
+            global_eds = get_global_editors() or None
+
+        installed_eds = install_ai_tools(target_dir, rules_dir, dry_run,
+                                         editors=global_eds)
         install_persona(target_dir, persona, dry_run)
         install_strict_git_hooks(profile, local, dry_run)
 
@@ -749,6 +766,10 @@ def main() -> None:
             extends_info=extends_info,
         )
 
+        # Record global editors (only for global install, not --local)
+        if not local and installed_eds:
+            record_global_editors(installed_eds)
+
         # Register project in global registry (for `ai-toolkit update` propagation)
         # Skipped when called from update_projects.py (--skip-register) to avoid
         # concurrent writes to projects.json during parallel updates.
@@ -756,10 +777,20 @@ def main() -> None:
             extends_source = ""
             if extends_info:
                 extends_source = extends_info.get("source", "")
+            # Determine editors to record for this project
+            local_eds_for_registry: list[str] | None = None
+            if local and local_editors_arg:
+                if local_editors_arg == "all":
+                    from install_steps.ai_tools import ALL_EDITORS
+                    local_eds_for_registry = list(ALL_EDITORS)
+                else:
+                    local_eds_for_registry = [e.strip() for e in local_editors_arg.split(",") if e.strip()]
+
             is_new = register_project(
                 project_dir,
                 profile=profile or "standard",
                 extends=extends_source,
+                editors=local_eds_for_registry,
             )
             if is_new:
                 print(f"  Registered project in {TOOLKIT_DATA_DIR / 'projects.json'}")