OE MCP Server

Connect your AI coding tools to enterprise data

One binary. One JSON config. No code. Give Claude Code, Cursor, Windsurf, and Claude Desktop direct access to your databases, files, and APIs via the Model Context Protocol.

Windows Linux macOS View on GitHub

45+ Connector Categories stdio + HTTP transport Persistent Memory Action Log Run OE Runtime Agents Self-Hosted · No Call-Home

🔌

45+ connector categories

PostgreSQL, MongoDB, S3, GitHub, Slack, Gmail, SSH, REST APIs, and more — all defined in a single oe-mcp.json file. No code required.

🧠

Persistent memory

Store facts across sessions with memory_set / memory_get. Every connector call is automatically logged to oe-mcp-log.json.

⚡

Run AI agents

Execute OE Runtime SKILL.md agents directly from Claude Code, Cursor, or any MCP client. Manual skills pause for your approval via approve_chain.

Quick Start

Up and running in 3 steps

No binary download needed — just npx.

Step 1 — Create oe-mcp.json

{
  "connectors": [
    {
      "name": "my-postgres",
      "type": "postgresql",
      "host": "localhost",
      "port": 5432,
      "database": "mydb",
      "user": "postgres",
      "password": "secret"
    }
  ]
}

Step 2 — Add to your AI app's MCP config

macOS / Linux

{
  "mcpServers": {
    "oe-mcp": {
      "type": "stdio",
      "command": "npx",
      "args": ["-y", "@openenthrium/oe-mcp", "--stdio", "/path/to/oe-mcp.json"]
    }
  }
}

Windows

{
  "mcpServers": {
    "oe-mcp": {
      "type": "stdio",
      "command": "npx.cmd",
      "args": ["-y", "@openenthrium/oe-mcp", "--stdio", "C:\\path\\to\\oe-mcp.json"]
    }
  }
}

⚠️ -y is required — without it npx blocks and the MCP connection never opens.

Step 3 — Reload your AI app

Connectors appear as tools automatically. Ask Claude: "What connectors do you have access to?" to verify.

Built-in Tools

Beyond connectors

OE MCP ships three built-in tool categories in addition to your connectors.

🧠 Persistent Memory

Key-value pairs that survive restarts — stored in oe-mcp-memory.json.

Tool	Description
memory_set	Store a key-value pair across sessions
memory_get	Retrieve a stored value by key
memory_list	List all stored key-value pairs
memory_delete	Remove a stored key

📋 Action Log

Every connector call is logged automatically to oe-mcp-log.json with timestamp, tool, input, and result.

Tool	Description
log_list	List recent connector calls — newest first, supports `limit`
log_clear	Clear all log entries

⚡ Run AI Agents

Execute OE Runtime SKILL.md agents directly from your AI coding tool — no terminal required.

Tool	Description
run_agent	Run an agent by file path. Auto skills execute immediately; manual skills pause and return `pending_skill_chain`.
list_pending_skills	List all manual skills currently paused and waiting for approval
approve_chain	Approve, skip, or abort a paused manual skill by `chain_id`

Example: "Run the OE Skills orchestrator and send a Slack message — skip anything you can't do, abort if it asks for credentials."

Transport Modes

Local or cloud — your choice

--stdio

stdio mode

Best for Claude Code, Cursor, Windsurf, Codex, Claude Desktop — launched as a child process by the AI app.

npx -y @openenthrium/oe-mcp --stdio oe-mcp.json

--serve

HTTP server mode

Best for cloud deployments or sharing one server across a team.

oe-mcp-linux --serve --port 4040 oe-mcp.json
# → http://your-server.com:4040/mcp

Sample Configs

Ready-to-use connector configs

Download oe-mcp-samples.zip — pre-built oe-mcp.json files for common connectors.

postgres mysql mongodb github slack gdrive ssh filesystem oracle salesforce servicenow telegram notion confluence graphql zoho-mail sftp dropbox multi-connector + more