Mcp Integration

Use Model Context Protocol servers to access external tools and data sources. Enable AI agents to discover and execute tools from configured MCP servers (legal databases, APIs, database connectors, weather services, etc.).

Install only if you trust every MCP server you configure and understand what its tools can read or change. Prefer HTTPS or local trusted servers, least-privilege and read-only credentials, per-agent/tool allowlists, and human approval for database writes, account changes, deletion, publishing, or administrative actions. Redact logs and configs before sharing diagnostics, and verify the MCP SDK dependency is updated.

Type: OpenClaw Skill Name: openclaw-mcp-plugin Version: 0.1.1 The OpenClaw MCP Integration plugin is designed to connect to external Model Context Protocol (MCP) servers and expose their tools to AI agents. The code (src/index.js, src/http-transport.js) implements the client-side logic for this, making network requests to configurable URLs. All documentation (SKILL.md, README.md, etc.) provides instructions for using the plugin and contains no evidence of prompt injection attempts, data exfiltration, or other malicious instructions directed at the AI agent. While connecting to external services inherently carries risks if those services are untrusted, the plugin itself does not exhibit intentional harmful behavior. The configuration schema (config/openclaw.plugin.json) mentions a 'stdio' transport with 'command' and 'env' fields, which would be a high-risk capability, but the current implementation (src/index.js) only supports 'http' transport, rendering these fields inert.