Tripitaka MCP

บริการพระไตรปิฎกบาลีผ่าน MCP protocol — เผยแผ่เป็นธรรมทาน

The Pāli Canon, accessible to your AI assistant. This MCP server lets Claude — or any MCP-compatible client — search, cite, and compare translations of the Buddha's discourses with segment-level precision. Offered freely as Dhamma Dāna for study, research, and practice.

Install in 1 click → or browse the canon · learn what it does ↓

What it does

The server exposes 12 tools and 3 resources over MCP. Your AI agent can:

• Search the canon by keyword (trigram), meaning (vector), or both (RRF hybrid)
• Fetch any sutta — in full, as a table of contents (mode="outline"), or just the segments around a search hit — with a cross-reference URL to the project's own bilingual reader. Even the longest sutta (DN 16, 1,664 segments) stays usable without flooding the context window
• Compare translations side-by-side, segment-aligned across editions
• Resolve Pāli words against P. A. Payutto's dictionary, PTS, and DPPN — including inflected forms
• Cite any line back to the bilingual reader with /read/{sutta_id}#{segment_id} — the page opens with the line highlighted and scroll-centered (try SN 56.11 segment 2.4)
• Open an interactive sutta viewer right inside the chat (MCP Apps) — Pāli and English side by side with the cited segment highlighted, plus an optional AI-translated third row in your own language, clearly badged and always verifiable against the originals. Works in MCP Apps-capable hosts such as Claude (web/desktop/mobile) and VS Code Copilot; other clients receive a regular text response

Coverage

Full Tipiṭaka in Pāli at parity with SuttaCentral's bilara-data (CC0). English from Bhikkhu Sujato (Sutta) and Bhikkhu Brahmali (Vinaya).

Use cases

Questions the server is built for — aggregation, comparison, and deep lookup across all ~444K segments.

Connect your AI

The hosted server is free and runs at:

https://mcp.tripitaka-mcp.com/mcp   (Streamable HTTP, MCP spec 2025-03-26)
https://mcp.tripitaka-mcp.com/sse   (legacy SSE, for older clients)

Quick install

For clients that support one-click install or simple paste:

No one-click for Claude Desktop yet — see the manual config below. Also listed on Glama for discovery.

Claude Desktop

You'll need Node.js installed (any recent version). The config below routes Claude Desktop to the remote server through the mcp-remote bridge.

1. Find your absolute npx path — Claude Desktop doesn't read your shell profile, so generic npx won't resolve. Open a terminal and run:

which npx
# example output: /Users/you/.nvm/versions/node/v22.x/bin/npx

2. Open claude_desktop_config.json (~/Library/Application Support/Claude/ on macOS, %APPDATA%\Claude\ on Windows) and add the entry below. Replace YOUR_NPX_PATH with the output from step 1, and YOUR_NODE_BIN_DIR with its parent directory:

{
  "mcpServers": {
    "tripitaka": {
      "command": "YOUR_NPX_PATH",
      "args": ["-y", "mcp-remote", "https://mcp.tripitaka-mcp.com/mcp"],
      "env": {
        "PATH": "YOUR_NODE_BIN_DIR:/usr/local/bin:/usr/bin:/bin"
      }
    }
  }
}

3. Quit Claude Desktop completely (⌘Q / right-click tray → Quit), then reopen. The 🔌 indicator in the bottom-left should show tripitaka connected with 12 tools available.

First connection takes 5–10 seconds while npx downloads mcp-remote on demand — give Claude Desktop a moment after restart before assuming it failed. The repo's annotated example config covers stdio (local install) and both remote transports.

Other MCP clients

• Cursor — add the same npx mcp-remote https://mcp.tripitaka-mcp.com/mcp entry to ~/.cursor/mcp.json (or the project-level .cursor/mcp.json). Restart Cursor; the tools appear under the MCP panel.
• Codex CLI & other Streamable-HTTP clients — point directly at https://mcp.tripitaka-mcp.com/mcp. No bridge needed.
• Legacy SSE clients — use https://mcp.tripitaka-mcp.com/sse instead.

The example config in the repo documents the full set of transports.

What to ask

See worked use cases with screenshots for six full examples — thematic surveys, translation comparisons, Pāli word studies, reading paths, and more. Claude picks the right tool automatically and links back to the bilingual reader for source verification.

Optional research skill

Drop skills/tipitaka-research.md into ~/.claude/skills/ to activate a multi-step research workflow (clarify → verify coverage → search → drill → cite) tailored to scriptural research.

Resources

• Source code — MIT-licensed, fork-friendly
• Topic pages — six curated markdown pages: canon overview, getting-started + tool selection, places (Mahājanapada + holy sites), themes (10 foundational teachings with locus classicus), people (~30 major figures), timeline (phase-based chronology of the 45-year mission). Sutta IDs verified against live data so AI agents can cite directly.
• NOTICE — data licenses (some components are non-commercial, please read before redistributing)