MCP server for pline.ai: scrape pages, crawl sites, batch-scrape URL lists, map a site's URLs, and search the web, all through your own pline.ai API key.

Set up every MCP client on your machine, plus the agent skill, in one step:

npx -y pline-api@latest init

It asks for your API key, registers the server with Claude Code, Codex, Cursor, VS Code, Claude Desktop, and Windsurf (merging and backing up existing configs), and installs the skill. Run npx -y pline-api@latest init --help for options.

Or point any MCP client at npx yourself:

{
  "mcpServers": {
    "pline.ai": {
      "command": "npx",
      "args": ["-y", "pline-api"],
      "env": {
        "PLINE_BASE_URL": "https://apix.pline.ai/v1",
        "PLINE_API_KEY": "YOUR-API-KEY"
      }
    }
  }
}

PLINE_API_KEY is your key, sent to the API as x-api-key. PLINE_BASE_URL is optional and defaults to the public pline.ai API, https://apix.pline.ai/v1.

The package is a small launcher. The server is a native binary selected for your platform (macOS arm64 and x64, Linux x64 and arm64, Windows x64) through an optional dependency, with a checksum-verified download from the GitHub Release as a fallback. Set PLINE_MCP_BINARY to run a build of your own. Arguments pass through: npx -y pline-api --http 127.0.0.1:8080 serves Streamable HTTP instead of stdio.

Tools: scrape, batch_scrape_start, batch_scrape_status, batch_scrape_cancel, crawl_start, crawl_status, crawl_cancel, map_site, search, sessions_list. Resources pline://guide and pline://reference document usage and every API field.

Client-specific setup (Cursor, VS Code, Claude Desktop, Claude Code, Windsurf, Docker, prebuilt binaries) and the agent skill are in the repository README. MCP Registry name: io.github.grepsr/pline-api.