GoPluto MCP — Up‑to‑date Service Knowledge & Snippets in Any Prompt

GoPluto MCP lets AI coding assistants pull fresh service/provider knowledge, routing rules, and ready‑to‑paste snippets from GoPluto — right inside your prompt.

❌ Without GoPluto

• ❌ No guarantee of finding the right expert when you actually need them
• ❌ Made-up providers and broken flows waste time
• ❌ Constant context-switching between people slows you down

✅ With GoPluto

Add use gopluto in your prompt (or set a rule once) and the assistant fetches live GoPluto context:

• Fresh, verified provider knowledge pulled directly into your prompt
• instant provider matching hints & filters
• minimal payloads & code snippets for expert access in minutes

Examples

Draft a WhatsApp onboarding flow that asks for experts within minutes. use gopluto

Show me the minimal payload to create a Service Request and start matching experts in under 60 seconds. use gopluto

🛠️ Installation

Requirements

• Any MCP‑capable client (Cursor, Claude Code, VS Code, Windsurf, Zed, etc.)
• GoPluto API Key (keep client‑side)
• Network access to https://api.gopluto.ai/mcp

Auth header: gopluto-api-key: YOUR_API_KEY

If your client can’t attach headers to remote servers, use the included local stdio proxy (main.py) to add the header for you.

Cursor

One‑click (Remote HTTP)

Manual (Remote)

{
  "mcpServers": {
    "gopluto": {
      "url": "https://api.gopluto.ai/mcp",
      "headers": { "gopluto-api-key": "YOUR_API_KEY" }
    }
  }
}

Local (stdio proxy)

{
  "mcpServers": {
    "gopluto": {
      "command": "uv",
      "args": ["run","python3","/ABSOLUTE/PATH/TO/gopluto-mcp/main.py"],
      "env": {
        "GOPLUTO_MCP_URL": "https://api.gopluto.ai/mcp",
        "GOPLUTO_AUTH_HEADER": "gopluto-api-key",
        "GOPLUTO_API_KEY": "YOUR_API_KEY",
        "GOPLUTO_TRANSPORT": "sse"
      }
    }
  }
}

Claude Code (CLI)

Remote

claude mcp add --transport http gopluto https://api.gopluto.ai/mcp \
  --header "gopluto-api-key: YOUR_API_KEY"

Local

claude mcp add gopluto -- uv run python3 /ABSOLUTE/PATH/TO/gopluto-mcp/main.py

Windsurf

Remote

{
  "mcpServers": {
    "gopluto": {
      "serverUrl": "https://api.gopluto.ai/mcp",
      "headers": { "gopluto-api-key": "YOUR_API_KEY" }
    }
  }
}

Local

{
  "mcpServers": {
    "gopluto": {
      "command": "uv",
      "args": ["run","python3","/ABSOLUTE/PATH/TO/gopluto-mcp/main.py"]
    }
  }
}

VS Code (Copilot Chat MCP)

Remote

"mcp": {
  "servers": {
    "gopluto": {
      "type": "http",
      "url": "https://api.gopluto.ai/mcp",
      "headers": { "gopluto-api-key": "YOUR_API_KEY" }
    }
  }
}

Local

"mcp": {
  "servers": {
    "gopluto": {
      "type": "stdio",
      "command": "uv",
      "args": ["run","python3","/ABSOLUTE/PATH/TO/gopluto-mcp/main.py"]
    }
  }
}

Cline

Marketplace → search GoPluto → Install

Manual (Remote)

{
  "mcpServers": {
    "gopluto": {
      "url": "https://api.gopluto.ai/mcp",
      "type": "streamableHttp",
      "headers": { "gopluto-api-key": "YOUR_API_KEY" }
    }
  }
}

Zed

{
  "context_servers": {
    "GoPluto": {
      "command": {
        "path": "uv",
        "args": ["run","python3","/ABSOLUTE/PATH/TO/gopluto-mcp/main.py"]
      }
    }
  }
}

Augment Code

UI → Settings → Tools → + Add MCP → Command:

uv run python3 /ABSOLUTE/PATH/TO/gopluto-mcp/main.py

Manual

"augment.advanced": {
  "mcpServers": [
    {
      "name": "gopluto",
      "command": "uv",
      "args": ["run","python3","/ABSOLUTE/PATH/TO/gopluto-mcp/main.py"]
    }
  ]
}

Roo Code

Remote

{
  "mcpServers": {
    "gopluto": {
      "type": "streamable-http",
      "url": "https://api.gopluto.ai/mcp",
      "headers": { "gopluto-api-key": "YOUR_API_KEY" }
    }
  }
}

Local

{
  "mcpServers": {
    "gopluto": {
      "command": "uv",
      "args": ["run","python3","/ABSOLUTE/PATH/TO/gopluto-mcp/main.py"]
    }
  }
}

Gemini CLI

Remote

{
  "mcpServers": {
    "gopluto": {
      "httpUrl": "https://api.gopluto.ai/mcp",
      "headers": {
        "gopluto-api-key": "YOUR_API_KEY",
        "Accept": "application/json, text/event-stream"
      }
    }
  }
}

Local

{
  "mcpServers": {
    "gopluto": {
      "command": "uv",
      "args": ["run","python3","/ABSOLUTE/PATH/TO/gopluto-mcp/main.py"]
    }
  }
}

Claude Desktop

Remote: Settings → Connectors → Add Custom Connector
URL: https://api.gopluto.ai/mcp
Header: gopluto-api-key: YOUR_API_KEY

Local: Developer settings → edit claude_desktop_config.json

{
  "mcpServers": {
    "gopluto": {
      "command": "uv",
      "args": ["run","python3","/ABSOLUTE/PATH/TO/gopluto-mcp/main.py"],
      "env": {
        "GOPLUTO_MCP_URL": "https://api.gopluto.ai/mcp",
        "GOPLUTO_AUTH_HEADER": "gopluto-api-key",
        "GOPLUTO_API_KEY": "YOUR_API_KEY",
        "GOPLUTO_TRANSPORT": "sse"
      }
    }
  }
}

Opencode

Remote

"mcp": {
  "gopluto": {
    "type": "remote",
    "url": "https://api.gopluto.ai/mcp",
    "headers": { "gopluto-api-key": "YOUR_API_KEY" },
    "enabled": true
  }
}

Local

"mcp": {
  "gopluto": {
    "type": "local",
    "command": ["uv","run","python3","/ABSOLUTE/PATH/TO/gopluto-mcp/main.py"],
    "enabled": true
  }
}

JetBrains AI Assistant

Settings → Tools → AI Assistant → Model Context Protocol (MCP) → + Add → As JSON

{
  "mcpServers": {
    "gopluto": {
      "command": "uv",
      "args": ["run","python3","/ABSOLUTE/PATH/TO/gopluto-mcp/main.py"]
    }
  }
}

Kiro

{
  "mcpServers": {
    "GoPluto": {
      "command": "uv",
      "args": ["run","python3","/ABSOLUTE/PATH/TO/gopluto-mcp/main.py"],
      "env": {},
      "disabled": false,
      "autoApprove": []
    }
  }
}

Trae

Remote

{
  "mcpServers": {
    "gopluto": { "url": "https://api.gopluto.ai/mcp" }
  }
}

Local

{
  "mcpServers": {
    "gopluto": {
      "command": "uv",
      "args": ["run","python3","/ABSOLUTE/PATH/TO/gopluto-mcp/main.py"]
    }
  }
}

Bun / Deno (local)

Bun

{
  "mcpServers": {
    "gopluto": {
      "command": "bunx",
      "args": ["uv","run","python3","/ABSOLUTE/PATH/TO/gopluto-mcp/main.py"]
    }
  }
}

Deno

{
  "mcpServers": {
    "gopluto": {
      "command": "deno",
      "args": ["run","--allow-run","--allow-env","--allow-net","/ABSOLUTE/PATH/TO/gopluto-mcp/deno_shim.ts"]
    }
  }
}

Docker (local HTTP)

Dockerfile

FROM python:3.11-slim
WORKDIR /app
COPY requirements.txt main.py ./
RUN pip install --no-cache-dir -r requirements.txt
ENV GOPLUTO_MCP_URL=https://api.gopluto.ai/mcp \
    GOPLUTO_AUTH_HEADER=gopluto-api-key \
    GOPLUTO_TRANSPORT=sse
EXPOSE 12006
CMD ["python","main.py","--mode","http","--host","0.0.0.0","--port","12006"]

Run:

docker build -t gopluto-mcp .
docker run --rm -p 12006:12006 -e GOPLUTO_API_KEY=YOUR_API_KEY gopluto-mcp

Point your client to http://localhost:12006.

Windows notes

If paths/quoting are tricky, use cmd /c or PowerShell; prefer absolute paths for command and args.

🔧 Available Tools (examples)

• resolve-gopluto-id – Resolve a package/service name to a GoPluto ID
• get-gopluto-docs – Fetch docs/snippets for a GoPluto ID
• search-providers – Query providers by skills/geo/budget
• create-service-request – Minimal payload helper

(Your upstream server’s tool list may evolve; use your client’s tool inspector to list them.)

🛟 Tips

Add a Rule to auto‑use GoPluto

Always use gopluto when I ask for service matching, onboarding flows, or API payloads.
Automatically call GoPluto MCP tools to resolve IDs and fetch snippets.

HTTPS Proxy: standard HTTPS_PROXY / https_proxy env variables apply.

💻 Development (local proxy)

uv pip install -r requirements.txt
export GOPLUTO_API_KEY="YOUR_REAL_KEY"
uv run python3 main.py               # stdio (default)
# or:
uv run python3 main.py --mode http --port 12006

Smoke test remote server

curl -X POST "https://api.gopluto.ai/mcp" \
  -H "Content-Type: application/json" \
  -H "gopluto-api-key: $GOPLUTO_API_KEY" \
  -d '{"jsonrpc":"2.0","id":"1","method":"initialize","params":{"protocolVersion":"2025-06-18"}}'

🚨 Troubleshooting

• 401/403 → wrong/missing gopluto-api-key
• Client can’t add headers → use local stdio proxy
• TLS issues → set SSL_CERT_FILE for Python
• Module not found (Node) → prefer Python proxy (no Node needed)

📄 License

MIT