Ai agents
Get AI agent
GET
Get one AI agent with its instructions, settings, skills and knowledge.
bearer agents:readheader
AuthorizationBearer <token>API key (rsl_live_…) from Settings → Developers.
Scope: agents:read
agentId*stringAI agent id. Never invent ids – list resources first.
Length
1 <= length <= 64instructions?stringfull = whole system prompt; summary = first 2,000 characters; none = omit it. Prompts can be 200k+ characters – use summary/none and agents_prompt_search when you only need part of it.
Default
"full"Value in
"full""summary""none"Success
application/json- response
id*stringname*stringdescription*stringisActive*booleanisDefaultWidgetAgent*booleanmodel*stringResponse model id; null = platform default.
activeDomains*array<string>dashboardUrl*stringcreatedAt*numberUnix time in milliseconds (UTC).
updatedAt*numberUnix time in milliseconds (UTC).
instructions*stringThe agent's system prompt (its main instructions); shortened or empty depending on the instructions parameter.
instructionsLength*numberFull prompt length in characters.
emailSignature*stringsettings*sidebar*skillIds*array<string>Enabled skills (system ids and custom_).
knowledgeSources*array<>knowledgeSize*curl -X GET "https://example.com/agents/string"{ "id": "string", "name": "string", "description": "string", "isActive": true, "isDefaultWidgetAgent": true, "model": "string", "activeDomains": [ "string" ], "dashboardUrl": "string", "createdAt": 0, "updatedAt": 0, "instructions": "string", "instructionsLength": 0, "emailSignature": "string", "settings": { "autoSummaryEnabled": true, "escalationReviewEnabled": true, "escalationReplyEnabled": true, "autoClose": { "enabled": true, "inactivityMinutes": 0 } }, "sidebar": { "label": "string", "emoji": "string" }, "skillIds": [ "string" ], "knowledgeSources": [ { "id": "string", "name": "string", "type": "string", "status": "string" } ], "knowledgeSize": { "chunkCount": 0, "contentBytes": 0 }}