Agents

Compose, run, and govern production agents over the API.

An Agent is a goal-directed reasoner that uses the Tasks, Skills, and Connections you make available to it. Collections are available through Task tools with RAG configured. See the concept guide. These endpoints let you manage agents and their runs programmatically: create and update agents, trigger runs, inspect run events and files, resolve pending approvals, and review revisions.

All endpoints are scoped to an organization and project and require a bearer token or OAuth client credentials. See Authentication for how to obtain and send credentials.

Key endpoints

POST
/api/v1/org/:org_id/project/:project_id/task-agent/:task_agent_id/run
1curl -X POST https://app.rightbrain.ai/api/v1/org/org_id/project/project_id/task-agent/task_agent_id/run \
2     -H "Authorization: Bearer <token>" \
3     -H "Content-Type: application/json" \
4     -d '{
5  "message": "Subject: Charged twice for June invoice\n\nHi, I think we'\''ve been billed twice for our June bookkeeping retainer — I can see two charges of £450 on the 3rd and the 5th. Could someone check our account and let me know what'\''s going on? Our account email is [email protected].\n\nThanks,\nDana"
6}'
Try it
MethodEndpointPurpose
POSTRun agentTrigger an agent run with your input; returns the result or streams events over SSE.
GETGet agent runFetch a single run’s status, output, timings, and cost.
POSTCreate agentCreate an agent with Tasks, Skills, and Connections.
POSTApprove requestResolve a human-in-the-loop pause; companion endpoints reject or resume a held run.
GETList revisionsInspect an agent’s versioned revisions to review or roll back changes.

The Agent Evals and Agent Shares endpoints (grouped in this section) cover offline evaluation runs and sharing an agent outside its project.