cc-skills: Instructions file for Claude Code

CLAUDE.md

This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.

Project Overview

This is a Claude Code plugin containing AI agent skills covering engineering, marketing, and productivity domains. The repository provides reusable skill definitions that Claude Code can invoke across projects.

Where content belongs: CLAUDE.md holds facts about this repository — structure, conventions, policy. Skills hold procedures and domain expertise for tasks Claude performs elsewhere. Never copy a CLAUDE.md fact into a skill body: the two drift, and a stale copy later contradicts the source with no signal which one is right.

Project Structure

skills/               # Claude Code skill definitions
  <skill-name>/
    SKILL.md          # Required: metadata + instructions
    references/       # Optional: detailed documentation loaded on demand
    scripts/          # Optional: executable code
    assets/           # Optional: templates, resources, config files, etc.
.claude-plugin/       # Plugin metadata and configuration
.cursor-plugin/       # Plugin metadata and configuration (version must match .claude-plugin/plugin.json)
gemini-extension.json # Gemini CLI extension manifest (version must match .claude-plugin/plugin.json)

Agent Skills Specification

All skills MUST conform to the Agent Skills specification. Key requirements are summarized below; the spec is the source of truth when in doubt.

Frontmatter

New skills go in skills/<skill-name>/SKILL.md. Each SKILL.md has YAML frontmatter. Fields per the Agent Skills spec — this project requires all fields marked "Project-required":

Field	Required	Constraints
name	Spec-required	1-64 chars. Lowercase a-z, digits, hyphens. No leading/trailing/consecutive hyphens. Must match parent directory name.
description	Spec-required	1-1000 chars (project hard cap; the AgentSkill spec allows 1024, but this project uses 1000 as a safety margin). Describes what the skill does and when to use it — this is the primary triggering mechanism. Be specific and slightly "pushy" to avoid under-triggering.
license	Project-required	License name or reference to a bundled license file. Use MIT for this project.
compatibility	Project-required	1-500 chars. Describe actual requirements as capabilities, never tool names — Requires internet access, not Requires WebSearch; the capability holds on every harness, the tool name only on Claude Code. Base: Designed for Claude Code, Codex or similar harness. Extend when needed: add Requires git, Requires internet access, Requires Python 3.14+ and uv, etc. Skills with no special requirements use the base string only.
metadata	Project-required	Must include author (string), version (semver a.b.c string, e.g. "1.0.0"), and openclaw (ClawHub discoverability fields — see below). Caution: some harnesses (e.g. OpenCode) parse metadata as a flat string→string map and may not preserve the nested openclaw object — don't assume every field survives outside Claude Code.
user-invocable	Project-required	Boolean. true for skills invocable as slash commands (e.g. /skill-name), false (default) for contextual skills that auto-trigger.
allowed-tools	Project-required	Space-delimited list of pre-approved tools. See "Allowed tools" below.
paths	Optional	Glob(s) scoping the skill to specific files/directories (e.g. **/*.go). Recognized by Cursor only — a no-op elsewhere. Add it for skills tied to one file type (framework or language skills) to sharpen triggering there; skip it for skills with no natural file-type scope.
dependencies	Optional, experimental	List of owner/repo@skill identifiers this skill should always load alongside. Formalizes an existing → See cross-reference as a machine-enforced co-load instead of prose the model might skip. Currently recognized by Antigravity only (third-party-documented, not yet confirmed in Google's official docs) — verify before relying on it, and keep the prose → See reference regardless since it's what every other harness actually reads.