Skip to main content

Overview

Retrieve detailed metadata, input fields, and output fields for any agent on the platform using its ID or slug.

Use Cases

  • Agent Discovery: Programmatically inspect an agent’s capabilities before invoking it.
  • Dynamic Integrations: Build tools that adapt to an agent’s input/output schema at runtime.
  • Catalog & Documentation: Generate agent directories or documentation from live metadata.

Endpoint

POST https://api-lr.agent.ai/v1/action/describe_agent

Authentication

Requires a Bearer token (your API key) in the Authorization header. Get your API key from agent.ai/user/settings#credits.

Configuration Fields

Agent ID

  • Description: The unique agent identifier or human-readable slug.
  • Example: "fluximage" or "1a4g81x0bfsc5dpi"
  • Required: Yes

Response Fields

Agent

  • id: Unique agent identifier.
  • human_id: Human-readable slug used in URLs (nullable).
  • name: Agent display name.
  • description: Agent description text.
  • url: Agent runner page URL (e.g. https://agent.ai/agent/fluximage).
  • profile_url: Agent profile page URL (e.g. https://agent.ai/profile/fluximage).
  • icon: Agent icon path.
  • type: Agent type — studio, workflow, or expert.
  • status: Visibility — public, private, or team.
  • tags: Array of category tags.
  • executions: Total number of runs.
  • review: Object with count (number of reviews) and score (average 1–5).
  • authors: Array of author objects with name, username, and user_token.
  • created_at: ISO 8601 creation timestamp.
  • updated_at: ISO 8601 last update timestamp.

Inputs

  • Description: Array of input field variable names the agent accepts.
  • Example: ["image_prompt"]

Outputs

  • Description: Array of output variable names the agent produces.
  • Example: ["image_generated"]

Example Request

-X POST https://api-lr.agent.ai/v1/action/describe_agent \

Example Response