Wexa documentation
Wexa gives AI workers governed access to what an organization knows (its context, knowledge, data and code) without copying any of it into a model. You need a Wexa project to follow any page here. The same tools are reachable four ways: the Wexa MCP server, the REST API, the TypeScript SDK and the Python SDK.
- Quickstart: get a key and make your first call.
- Connect an AI client: add Wexa to Claude, Cursor or ChatGPT.
- Core ideas: organizations, projects, scope and the four surfaces.
- Authentication: pick, create and revoke a credential.
- Errors and retries: refusals, typed errors and retrying.
- Changelog: what changed, and in which SDK version.
Next: Build an agent, connect your data, or read How every tool call works.
Check what your deployment has
Not every deployment carries every tool, and a missing tool is missing on all four surfaces. GET /v1/connection-info lists what yours has. For a missing tool, a refused call or a retired endpoint, see Troubleshooting.
Make a call now
Set WEXA_WORKSPACE and WEXA_API_KEY, then call the docs tool, which reports your own scope and limits. Quickstart covers getting the key.
curl -sS -X POST https://fabric.wexa.ai/v1/docs \
-H "Authorization: Bearer $WEXA_API_KEY" \
-H "Content-Type: application/json" \
-d '{"topic": "governance"}'{ "lifecycle_id": "qlc_...", "result": { "mode": "auto", "quota": { "project_window_max": 120 } } }