LAP CLI compiles API specifications (OpenAPI, GraphQL, AsyncAPI, Protobuf, Postman) into a lean, agent-native format (.lap), lets you search a registry of published specs, and can generate or install skills from those specs for AI agents. It streamlines discovering, acquiring, and publishing API specs so agents work with accurate, compressed API descriptions.
Use LAP when you need to find an API for an agent to call, convert local API specs into the LAP format for agent consumption, generate installable skills from a spec, or publish compiled specs to the LAP registry. Helpful during development, integration testing, or when building agent workflows that depend on precise API contracts.
Primarily CLI-oriented agents and agent environments that support external binaries or npx installs (Claude Code, Cursor, Copilot CLI, general LLM-driven CLIs). The skill notes using Claude CLI for AI-enhanced generation but supports local/npx fallbacks.
LAP CLI is a skill for compiling, searching, and managing API specifications (OpenAPI, GraphQL, AsyncAPI, etc.) into a compact agent-native .lap format. No bundled scripts — purely a command-driven SKILL.md with detailed command reference, error recovery, and two workflows (consumer and publisher). Well-structured with clear progressive disclosure via references/. Minor security note: instructs npm/npx installs without checksum verification.
Clean skill, no scripts to audit statically. Command-only skill with thorough documentation. The $LAP_CMD resolution pattern is a good practice for binary portability.