What problem does it solve?
It solves the problem of turning an MCP server or an existing SKILL.md skill into a standalone, runnable command-line interface so users can operate without relying on a live AI runtime or MCP connection.
Core Features & Use Cases
- Compile MCP backends into Click CLIs: Register an MCP server, discover its tools, and convert tool schemas into deterministic CLI commands.
- Compile SKILL.md into just-bash or legacy Python CLIs: Transform the skill’s steps and API endpoints into executable commands, with a TypeScript default output for the just-bash runtime.
- Batch compilation and selective skipping: Compile multiple skills from a directory while automatically skipping skills that lack endpoints or are marked un-compilable.
Use case: you have a tool-backed workflow exposed via MCP (or a skill catalog defined by SKILL.md) and you want a single set of CLI commands that can be run locally or in CI with repeatable parameters.
Quick Start
Compile a skill into a standalone CLI by asking: "Compile this SKILL.md into a runnable CLI using LiteCLI."