Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
Version History
V1.1: Added semantic search, hybrid scoring, and AI-powered task matching. Uses local char n-gram embeddings (384d). Hybrid search combines keyword (BM25-style) and semantic scores 50/50. New suggest command with confidence ratings (High/Medium/Low). Enhanced keyword scoring: exact match +20, name contains +10, triggers +3.
What This Skill Does
Searches available skills by keyword, semantic similarity, or natural language task description. A local index of skill metadata and pre-computed embeddings powers all three modes. Indexing happens automatically on first semantic search.
Three search modes in one script cover exact names, conceptual matches, and task descriptions, so the right skill can be found without knowing its exact name.
When to use it
- Discovering what audio transcription skills are installed
- Finding the right skill before starting a PDF workflow
- Checking if an image generation skill exists by describing the task
- Listing all available skills in the current environment
- Matching a vague task description to the most relevant skill
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: What skill should I use for transcribing audio?
- 1Runs LLM suggest mode: skill_search.py suggest "transcribe audio"
- 2Receives ranked skill recommendations with confidence scores and reasoning
- 3Reads SKILL.md for the top match to confirm it fits the task
- 4Loads the matched skill and proceeds with execution
Recommended skill: openai-whisper-api with High confidence. Reason: task explicitly matches skill purpose.