@loopyc/mcp
v0.9.0
Published
loopc-mcp — Model Context Protocol server exposing the Monkey D Loopy factory to agents (schema, scaffold, validate, verify, compile, run, inspect).
Maintainers
Readme
@loopyc/mcp
loopc-mcp — the whole Monkey D Loopy factory
exposed to agents as MCP tools.
Register the server and any MCP-capable agent can scaffold, validate, dry-run-verify, score,
compile, and inspect bounded agent loops — the same pipeline as the loopc CLI, tool by tool.
{ "mcpServers": { "loopy": { "command": "npx", "args": ["--yes", "@loopyc/mcp@latest"] } } }codex mcp add loopy -- npx --yes @loopyc/mcp@latest
claude mcp add --scope user loopy -- npx --yes @loopyc/mcp@latestThe server uses stdio. A global npm i -g @loopyc/mcp install and loopc-mcp command work too.
Tool reference: docs/mcp.md.
Jev authoring and iterative refinement (0.9.0)
These capabilities are available starting with 0.9.0. See availability and source setup and confirm that the installed CLI/MCP tool list includes the feature before using it.
| CLI | MCP | Purpose |
|---|---|---|
| recommend | recommend_workflow | Rank eligible catalog structures for a brief |
| design | design_workflow | Create a validated, mock-verified scaffold from an explicit selection |
| refine | refine_workflow | Compare current YAML and authored revisions with feedback/evidence; preserve history |
Jev requires TYPESAFE_API_KEY in the process environment and explicit provider selection
(allowExternal: true in MCP). Env files are not loaded automatically. Refinement sends supplied
eligible YAML, feedback and evidence to TypeSafe; keep secrets out. Offline performs local checks
and retains current, without claiming semantic improvement. These authoring calls never activate
a workflow or execute its real effects.
Read the workflow designer, agent guide, and complete agent context for exact contracts, setup, repeated rounds and limitations.
