Skip to content

Commit 59875ca

Browse files
sjnimsclaude
andcommitted
docs(CLAUDE.md): enumerate skills and document allowed-tools
- Add explicit list of all 7 skills with brief descriptions - Clarify bundled resource directory purposes (references vs examples vs scripts) - Document optional allowed-tools frontmatter for skills/agents - Mark command description as required field 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
1 parent 7d2034a commit 59875ca

File tree

1 file changed

+26
-6
lines changed

1 file changed

+26
-6
lines changed

CLAUDE.md

Lines changed: 26 additions & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -42,9 +42,19 @@ When testing locally, point to the plugin directory, not the root.
4242
Each skill follows progressive disclosure:
4343

4444
- `SKILL.md` - Core content (1,500-2,000 words, lean)
45-
- `references/` - Detailed documentation (loaded as needed)
46-
- `examples/` - Working code examples
47-
- `scripts/` - Utility scripts
45+
- `references/` - Detailed documentation loaded into context as needed
46+
- `examples/` - Complete working examples and templates for copy-paste
47+
- `scripts/` - Utility scripts (executable without loading into context)
48+
49+
**The 7 Skills:**
50+
51+
1. `hook-development` - Event-driven automation with prompt-based hooks
52+
2. `mcp-integration` - Model Context Protocol server configuration
53+
3. `plugin-structure` - Directory layout and manifest configuration
54+
4. `plugin-settings` - Configuration via .claude/plugin-name.local.md files
55+
5. `command-development` - Slash commands with frontmatter
56+
6. `agent-development` - Autonomous agents with AI-assisted generation
57+
7. `skill-development` - Creating skills with progressive disclosure
4858

4959
### Writing Style
5060

@@ -134,9 +144,19 @@ Skills require:
134144

135145
Commands are markdown files with frontmatter:
136146

137-
- `description`: Brief explanation
138-
- `argument-hint`: Optional argument placeholder
139-
- `allowed-tools`: Array of permitted tools
147+
- `description`: Brief explanation (required)
148+
- `argument-hint`: Optional argument placeholder text
149+
- `allowed-tools`: Optional array of permitted tools (restricts tool access)
150+
151+
### Skills/Agents Optional Frontmatter
152+
153+
Both skills and agents support `allowed-tools` to restrict tool access:
154+
155+
```yaml
156+
allowed-tools: Read, Grep, Glob # Read-only skill
157+
```
158+
159+
Use for read-only workflows, security-sensitive tasks, or limited-scope operations.
140160
141161
### Hooks
142162

0 commit comments

Comments
 (0)