OpenClaw integration

March 28, 2026 ยท View on GitHub

TicketLens can be integrated into OpenClaw most easily as a workspace skill that calls the public TicketLens REST API.

Option 1: workspace skill via the public API

This is the recommended OpenClaw path because it works across OpenClaw backends and follows OpenClaw's documented skill model directly.

Setup:

  1. Copy examples/openclaw/skills/ticketlens-api/ into your OpenClaw workspace under skills/ticketlens-api/, or into ~/.openclaw/skills/ticketlens-api/ for a shared local install.
  2. Start a new OpenClaw session, or refresh skills if your setup watches the workspace skill directory.
  3. Ask for destination experiences in natural language, for example:
    • Find Eiffel Tower tickets in Paris this weekend
    • Search hop-on hop-off buses in Barcelona
    • Look up football tickets in London next Friday

Notes:

  • This skill assumes the agent can use a runtime shell tool such as exec or bash.
  • The included skill uses the public TicketLens API with curl, so curl must be available on the OpenClaw host.
  • If the user names a landmark or attraction, the skill resolves the POI first and then searches experiences against that POI.

MCP note

OpenClaw's bundle docs currently describe MCP support around supported stdio servers launched as subprocesses. The public TicketLens MCP server is hosted remotely at https://mcp.ticketlens.com/, so the API skill above is the clearest "works today" integration path.

If you want MCP semantics inside OpenClaw, the more advanced path is to build either:

  • a native OpenClaw plugin that exposes TicketLens tools directly, or
  • a local stdio bridge that proxies to the hosted TicketLens MCP endpoint.

OpenClaw's ACP bridge docs also say per-session mcpServers are unsupported, so this is not a per-chat MCP setup path.

Relevant OpenClaw docs