Agent_skills agent-capability-discovery

name: agent-capability-discovery version: 1.0.0 description: Scans all skill directories in the repository to generate a comprehensive global map of agent capabilities, inputs, and outputs. Use when you need to understand the full potential of your agent library or when a master agent needs to decide which sub-agent skill to invoke for a complex task. inputs: base_directory: type: string description: The root directory containing all agent skill folders. default: "." output_format: type: string enum: [json, markdown, yaml] default: markdown outputs: capabilities_map: type: string description: A structured mapping of all discovered skills and their primary functions. capabilities:

• Recursive directory traversal to identify skill folders.
• Extraction of 'capabilities' and 'description' from skill.yaml.
• Categorization of skills based on their name or metadata.
• Generation of a routing table for agentic decision-making. constraints:
• Only scans directories containing a valid skill.yaml file.
• Accuracy depends on the clarity of descriptions in the target skills. security:
• Reads metadata only; no execution of third-party code.
• Operates strictly within the provided base_directory. examples:
• input: base_directory: "./agent-skills" output_format: "markdown" output: capabilities_map: "| Skill | Capability |\n| --- | --- |\n| secret-leak-detector | Scans code for secrets |"