// docs / mcp
Your library, inside your agents.
The MCP server lets an agent read, search, and fill your prompts directly. It's read-only by design, so it can never burn your tokens or change your library.
Connect it to Claude Code
# sign in with the CLI first (MCP shares the same session)
npm i -g promptlike && ple login
# add it to Claude Code
claude mcp add promptlike -- npx -y promptlike-mcpThe package is live on npm as promptlike-mcp. Zero extra auth: the server reads the same ~/.config/promptlike/config.json session the CLI wrote. Any MCP client works: point it at npx -y promptlike-mcp over stdio.
The tools (v1, read-only)
list_prompts(tag?) → browse your library, incl. {{var}} names
search_prompts(query) → search titles & descriptions
get_prompt(name, version?, project?) → full content, optionally a pinned version
fill_variables(name, vars, project?) → fill {{var}}s, return ready-to-use textRead-only is intentional: an agent can browse and fill your prompts, but it can never spend your AI tokens or mutate the library. Ambiguous names return a candidate list (each with its repo) so the agent picks, and it can retry with project to scope to one repo; rate limits surface as Retry in Ns so it can back off.
Why this matters
MCP has a first-class prompts concept but no versioning, diff, or history; that stays in PLE. So agents pull the current version over MCP while you keep the full engineering history on the web. You version it; your agents read it. More on the approach in variables and versioning.