CLS MCP Server (v1.0.3) connects LLMs (e.g., Claude, GPT) to Tencent Cloud Log Service via the Model Context Protocol. Provides 17+ tools for log search (CQL), metric queries (PromQL), alarm management, time conversion, and region lookup. Requires Tencent Cloud API credentials. Active development on GitHub (Tencent), supports stdio and SSE transports, Node.js 18–22. Distinct from generic MCP servers by deep CLS integration, natural language query translation, and built-in alarm workflow tools.
npm install cls-mcp-serverVerified import paths — ran on the pinned version, not inferred.
Initializes and runs the CLS MCP server with stdio transport, reading Tencent Cloud credentials from environment variables and setting region, max length, and timezone.
Ensure TRANSPORT is set to 'stdio' or 'sse' in environment, or pass in options if supported.
Use Node.js 18, 20, or 22.
Rename parameter to 'maxLength'.
Set TENCENTCLOUD_SECRET_ID and TENCENTCLOUD_SECRET_KEY environment variables.
Use the new prefixed variable names.
Set TENCENTCLOUD_SECRET_ID and TENCENTCLOUD_SECRET_KEY in the environment.
Use 'const runServer = require('cls-mcp-server').default' for CommonJS.Set TRANSPORT to 'stdio' or 'sse'.
Use the default 'tencentcloudapi.com' when not in Tencent Cloud VPC.
No dependency data recorded yet.