SkillCheck validator, Cursor exports, and per-agent installers (#27)
Three more learnings from alirezarezvani/claude-skills, applied: 1. SkillCheck validator (scripts/skillcheck.mjs) — validates every SKILL.md against the authoring standard (frontmatter, name/folder match, trigger + produces clauses, required headings) plus tier referential integrity. Errors fail CI; --strict fails on warnings too. New skillcheck.yml workflow and a SkillCheck status badge in the README. Current: 0 errors / 14 advisory warnings across 172 skills. 2. Cursor export platform — build-exports.mjs now generates exports/cursor/<bundle>/<skill>/<skill>.mdc rule files. The PLATFORMS registry now supports per-skill filenames (file as a function). 3. Per-agent installers — scripts/install.sh unifies install for claude/hermes/codex/openclaw/cursor (--link, --target, --dry-run, --list). Curl-able one-liners codex-install.sh, openclaw-install.sh, and cursor-install.sh clone the library and install in a single command. README documents the one-line installs and Cursor exports; CHANGELOG and the authoring standard updated. Claude-Session: https://claude.ai/code/session_016JWn5jRD5tcEFKrubjQ6Px Co-authored-by: Claude <noreply@anthropic.com>
This commit is contained in:
@@ -3,6 +3,7 @@
|
||||
[](https://github.com/mohitagw15856/pm-claude-skills/stargazers)
|
||||
[](https://github.com/mohitagw15856/pm-claude-skills)
|
||||
[](#-works-with--cross-tool-compatibility)
|
||||
[](.github/workflows/skillcheck.yml)
|
||||
[](https://github.com/mohitagw15856/pm-claude-skills/releases)
|
||||
[](https://github.com/mohitagw15856/pm-claude-skills#-quick-install-2-minutes)
|
||||
[](LICENSE)
|
||||
@@ -93,8 +94,10 @@ body as a system prompt — for those we ship ready-made [exports](#ready-to-use
|
||||
|---|---|---|
|
||||
| **Claude Code** (CLI / desktop / web / IDE) | Native. Install via the plugin marketplace; Claude loads a skill automatically when your request matches its description. | ✅ Yes |
|
||||
| **Hermes Agent** (Nous Research) | Native — same open `SKILL.md` standard. Run `python3 scripts/sync-hermes-skills.py` to install into `~/.hermes/skills/`; Hermes auto-discovers them. | ✅ Yes |
|
||||
| **OpenAI Codex · OpenClaw** | Native `SKILL.md`. One-line install (see [below](#one-line-install-for-coding-agents)) or `./scripts/install.sh --agent codex`. | ✅ Yes |
|
||||
| **Cursor** | Generated `.mdc` rules. One-line install into `.cursor/rules/`, or copy from [`exports/cursor/`](exports/cursor/). | ⚙️ By description |
|
||||
| **Claude.ai & Claude API** | Upload a skill, or paste the body in as a system prompt / project instruction. | ⚙️ Manual |
|
||||
| **Other coding agents that read the `SKILL.md` format** (e.g. Codex, Gemini CLI, Cursor) | Point the agent at the `skills/` folder, or paste the body. The frameworks are tool-agnostic; only the auto-discovery mechanism differs per tool. | ⚙️ Varies by tool |
|
||||
| **Other `SKILL.md`-aware agents** (Gemini CLI, Aider, Windsurf, …) | Point the agent at `skills/`, or use `./scripts/install.sh`. The frameworks are tool-agnostic; only discovery differs per tool. | ⚙️ Varies by tool |
|
||||
| **ChatGPT & Gemini** | Copy a ready-made export (below) into a Custom GPT or Gem's instructions. You keep the full framework and output format. | ❌ Paste per use |
|
||||
|
||||
**What's verified vs. what varies:** the skill **bodies** — the frameworks, rubrics, and
|
||||
@@ -111,16 +114,33 @@ maintained twice:
|
||||
|
||||
- **ChatGPT** — copy any [`exports/chatgpt/<bundle>/<skill>/SYSTEM_PROMPT.md`](exports/chatgpt/) straight into a Custom GPT's instructions.
|
||||
- **Google Gemini** — copy any [`exports/gemini/<bundle>/<skill>/GEM_INSTRUCTIONS.md`](exports/gemini/) into a Gem's instructions.
|
||||
- **Cursor** — copy any [`exports/cursor/<bundle>/<skill>/<skill>.mdc`](exports/cursor/) into `.cursor/rules/` (or use the one-liner below).
|
||||
|
||||
Native `SKILL.md` agents don't need exports — install the skills directly:
|
||||
### One-line install for coding agents
|
||||
|
||||
Native `SKILL.md` agents (Claude Code, Hermes, Codex, OpenClaw) and Cursor can install every skill with a single command — each clones the library and drops the skills where the agent discovers them:
|
||||
|
||||
```bash
|
||||
# Hermes Agent (Nous Research) — installs skills/ into ~/.hermes/skills/
|
||||
python3 scripts/sync-hermes-skills.py # copy (use --link to symlink, --dry-run to preview)
|
||||
# OpenAI Codex
|
||||
bash <(curl -fsSL https://raw.githubusercontent.com/mohitagw15856/pm-claude-skills/main/scripts/codex-install.sh)
|
||||
|
||||
# OpenClaw
|
||||
bash <(curl -fsSL https://raw.githubusercontent.com/mohitagw15856/pm-claude-skills/main/scripts/openclaw-install.sh)
|
||||
|
||||
# Cursor (.mdc rules into ./.cursor/rules)
|
||||
bash <(curl -fsSL https://raw.githubusercontent.com/mohitagw15856/pm-claude-skills/main/scripts/cursor-install.sh)
|
||||
```
|
||||
|
||||
Already cloned? Use the unified installer for any agent (add `--link`, `--target`, or `--dry-run`):
|
||||
|
||||
```bash
|
||||
./scripts/install.sh --list # claude · hermes · codex · openclaw · cursor
|
||||
./scripts/install.sh --agent codex # install for a specific agent
|
||||
python3 scripts/sync-hermes-skills.py # Hermes equivalent (copy, or --link / --dry-run)
|
||||
```
|
||||
|
||||
The skill body in `skills/<name>/SKILL.md` is the single source of truth. Regenerate the
|
||||
chat-LLM exports (or add a new platform — it's a few lines in the `PLATFORMS` registry) with:
|
||||
chat-LLM / Cursor exports (or add a new platform — it's a few lines in the `PLATFORMS` registry) with:
|
||||
|
||||
```bash
|
||||
node scripts/build-exports.mjs # regenerate all platform exports
|
||||
|
||||
Reference in New Issue
Block a user