openclaw-skills

Personal collection of OpenClaw skills for en's agent(s). Each skill is a self-contained directory with a SKILL.md (manifest), optional config.yml, scripts/, and assets/.

Skills

Skill	Purpose	Stack
comfyui-api	Low-level ComfyUI HTTP API client — generation, upscaling, img2img editing. Returns structured JSON for downstream consumers.	Python (stdlib only)
comfyui-image	Channel-aware wrapper that picks generate/upscale/edit, calls comfyui-api, and returns results to the originating surface (Eden / Discord / Matrix) with explicit fallback.	Policy / dispatch
nc-upload	Minimal Nextcloud upload + share-link helper for binary files (esp. images). Reads creds from ~/.openclaw/nextcloud/.env. Binary-safe (fixes upstream openclaw-nextcloud utf-8 read bug).	Node.js 24+ (stdlib only)

Conventions

• Every skill has a SKILL.md with at minimum:
  • name, description, read_when, metadata.emoji
  • "Scope" / "Usage" / "Safety" sections
• No required environment variables for install-time — runtime env loaded via .env or process env
• Scripts stay stdlib-only (no pip install or npm install for end users)
• Each skill is independently removable

Install

These are personal skills, but the layout follows the standard OpenClaw workspace-skill shape. Copy any skill directory to:

• Workspace skill: ~/.openclaw/workspace/skills/<name>/ (per-agent)
• Or use openclaw skills install <name> once they're indexed in ClawHub

Changelog

See CHANGELOG.md for version history of every skill.

Contributing

This is a personal repo. PRs from en only.

License

Private — not for redistribution.