cc-config
Use when working with Claude Code CLI, plugins, hooks, MCP servers, skills, configuration, or any Claude Code feature - routes to the official documentation instead of guessing, and can cache it locally for offline use
- License 许可
- Apache-2.0
- 所属仓库
- zak-keown/moe
- 最近更新
- 2026-09-05
- Stars
- ★ 0
- Forks
- 1
01 前言 Front Matter
- name
- cc-config
- description
- Use when working with Claude Code CLI, plugins, hooks, MCP servers, skills, configuration, or any Claude Code feature - routes to the official documentation instead of guessing, and can cache it locally for offline use
02 资源文件 2
- 脚本scripts/update_docs.d.mts160 B
- 脚本scripts/update_docs.mjs4.2 KB
03 SKILL.md 正文
Working with Claude Code
Overview
Claude Code's own documentation is the authority on configuration paths, manifest schemas, hook events and CLI flags. This skill's job is to send you there rather than let you guess — and, when you want it offline, to fetch it.
It ships no documentation of its own. The upstream version of this skill
committed a 42-file mirror of docs.claude.com/en/docs/claude-code/*. That copy
was dropped on import: it was nine months stale, it omitted eight plugin manifest
keys that this monorepo's own packages/mint/schemas/claude-code-plugin-manifest.json
already validates, and a skill whose pitch is "authoritative, stop guessing" that
is behind the repo's own schema launders a wrong answer as an official one.
When to Use
- Creating or configuring Claude Code plugins
- Setting up MCP servers
- Working with hooks (SessionStart, Stop, PostToolUse, …)
- Writing or testing skills
- Configuring Claude Code settings
- Troubleshooting Claude Code issues
- Understanding CLI commands
- Setting up integrations (VS Code, JetBrains, …)
- Configuring networking, security, or enterprise features
How to Get an Answer
Default: fetch the one page you need. Every topic is a single markdown file under one predictable URL:
https://docs.claude.com/en/docs/claude-code/<topic>.md
Use webfetch on that URL. One page, current, no cache to go stale.
| Question | Topic |
|---|---|
| Create a plugin | plugins, then plugins-reference |
| Plugin marketplaces | plugin-marketplaces |
| Set up an MCP server | mcp |
| Configure hooks | hooks, then hooks-guide |
| Write a skill | skills |
| Slash commands | slash-commands |
| Subagents | sub-agents |
| CLI flags | cli-reference |
| Headless / scripting | headless |
| Settings reference | settings |
| Memory and context | memory |
| Costs and usage | costs, monitoring-usage |
| Troubleshoot | troubleshooting |
| Install and set up | setup, quickstart |
https://docs.claude.com/llms.txt lists every page if the topic you want is not
in the table.
Working Offline
When you have no network, or you want to search across the whole corpus with
grep, populate a local cache first:
Resolve skills/cc-config/scripts/update_docs.mjs
relative to this loaded document, then invoke it with node.
It reads llms.txt, extracts every claude-code/*.md URL, and writes each page
into the sibling references/ directory beside the resolved script's parent
directory. Then use read or grep on that directory as usual.
Two things to know about the cache:
- It is not committed. A fresh checkout has no
references/directory at all, and this skill's answers do not depend on one existing. - It goes stale silently. Nothing dates it and nothing warns you. If a cached page disagrees with a fetched one, the fetched one wins — re-run the script rather than reasoning from the cache.
What This Skill Does NOT Do
- It provides documentation access, not procedural guidance. For how to
build a plugin — the workflow, the manifest, the release path in this
repo — use the
cc-pluginsskill. - It is a reference router, not a tutorial.
Red Flags
If you find yourself:
- Guessing where a config file lives → fetch
settings - Speculating about a manifest key → fetch
plugins-reference - Unsure which hook event fires when → fetch
hooks - Assuming a feature exists → check the docs first
Consult the documentation before guessing. Do not answer from a stale cache when the network is available.
04 同仓库更多 Skills
write-skill
Use when creating new skills, editing existing skills, or verifying skills work before deployment
write-plan
Use when you have a spec or requirements for a multi-step task, before touching code
write-clearly
Apply Strunk's timeless writing rules to ANY prose humans will read—documentation, commit messages, error messages, explanations, reports, or UI text. Makes your writing clearer, stronger, and more professional.
windows-vm
Create, manage, or connect to a headless Windows 11 VM running in Docker with SSH access. Use when the user wants to spin up, stop, restart, or SSH into a Windows VM.
verify-completion
Use when about to claim work is complete, fixed, or passing, before committing or creating MRs - requires running verification commands and confirming output before making any success claims; evidence before assertions always
using-moe
Use when starting any conversation - establishes how to find and use skills, requiring skill invocation before ANY response including clarifying questions