Create agent
Create a new agent. The authenticated caller must have permission to create agents on the target team.
Authorizations
Body
Display name for the agent.
"Sales research agent"
ID of the LLM the agent runs on. Use GET /models to discover valid values. Omit for the organization's default agent model.
"anthropic/claude-sonnet-4"
"Researches accounts and drafts outreach"
"You are a B2B sales research assistant."
Connectors and native abilities to start with. The platform's default native abilities are added behind whatever you send. Prefer PUT /agents/{agent_id}/mcp-servers/{server_id} and PATCH /agents/{agent_id}/abilities after creation.
Resources attached to the agent.
IDs of skills to attach to the agent. Attachment happens inside the create transaction, so an invalid ID fails the whole request (no orphaned agent). Omit to attach none. The caller must hold INVOKE on each skill. After creation, manage skills with PATCH /agents/{agent_id}/skills.
Brain sources to attach, with an optional scope each. Attached inside the create transaction, like skill_ids. Omit config for the whole source.
Agent settings. Omitted sections take the organization's agent defaults.
ID of the folder to place the agent in.
"folder_91ab"
Whether the agent is active. Defaults to true.
Setting this to false retires the agent: it disappears from GET /agents, and GET/PATCH /agents/{agent_id} return 404, so it cannot be reactivated through the API. This is not a pause switch — to stop an agent from running while keeping it reachable, disable its triggers instead.
Optional caller-supplied agent ID. When omitted, the server generates one.
ID of the team to create the agent under. When omitted, the agent is owned by the authenticated user.
"team_4f8c92ab"
Response
Agent created.
An agent and its whole configuration. Single-agent retrieve, create and update return this shape; the list endpoint does not populate the inlined collections, abilities or version. Collection-specific writes return their own result shapes.