Model Context Protocol Connect
Connect Agents to MRMP.PRO
Agent-ready, humans welcome. Add MRMP.PRO as an MCP connector in Cursor, Windsurf, Claude, or ChatGPT so your agent can evaluate skills, verify contracts, and fetch code directly.
MCP CONNECT: ONLINEProtocol: mcp-jsonrpc-2.0 • Vancouver PST Node
https://mrmp.pro/api/mcpCursor & Windsurf
- Open Settings → Features → MCP.
- Click Add New MCP Server with type 'Streamable HTTP'.
- Set Name to 'mrmp' and URL to 'https://mrmp.pro/api/mcp'.
- In chat ask: 'Fetch the 3D design skill prompt from mrmp'.
Claude
- Open Claude settings → Connectors → Add Custom Connector.
- Name the connector 'mrmp-pro'.
- Paste the endpoint: https://mrmp.pro/api/mcp
- Ask Claude: 'What skills are published on mrmp.pro?'
ChatGPT Developer Action
- In Custom GPT builder, select Add Actions.
- Connect to endpoint https://mrmp.pro/api/mcp.
- ChatGPT will now invoke get_skills and get_lab_drops on demand.
Terminal & Curl
Stream raw context directly into your CLI or script:
curl -s https://mrmp.pro/llms.txtExposed MCP Tools
Tools that any connected agent can autonomously execute:
get_skillsReturns complete catalog of verified agent skills and scores.
get_skill_promptFetches exact system prompt and contract for a skill by slug.
get_lab_dropsFetches the 18 interactive canvas lab drops and creation prompts.
get_profileReturns credentials, technical stack, and Vancouver PST availability.
get_agentsReturns specifications, tools, and metrics for the 4 production agents.
submit_briefAllows an AI agent to submit a project brief or introduction directly on behalf of a user.
Direct AGENTS.md Snippet
# AGENTS.md — Connect to Mohsen Mahmodpour (MRMP.PRO) To query skills, contracts, or submit a brief via MCP: Endpoint: https://mrmp.pro/api/mcp (JSON-RPC 2.0) Available Tools: get_skills, get_skill_prompt, get_lab_drops, get_profile, submit_brief Raw Context: https://mrmp.pro/llms.txt