Get the active AI prompt (internal)

Add MCP server to your AI tool

Allow AI tools and LLMs to interact with the API documentation portal through MCP.

MCP server URL


        
          
            https://docs.usehearsay.com/mcp

Standard setup for AI tools providing an mcp.json file

{
  "integrations MCP server": {
    "url": "https://docs.usehearsay.com/mcp"
  }
}

GET /ai/admin/prompts/{purpose}/active

Access token

Returns the prompt the pipeline uses for one purpose: the active stored version, or the code default when none is stored. The result is cached for up to a minute and has a different shape from a stored version (the tool fields are grouped under tool). An unknown purpose returns 400. Super admins only.

Path parameters

purpose string Required

Values are chunk_summary, period_summary, conversation_summary, case_summary, case_context_filing_extraction, relevance_gate, evidentiary_map_element, or tag_suggestion.

Responses

200 application/json
Hide response attributes Show response attributes object
- data object Required
  
  Hide data attributes Show data attributes object
  
  purpose string
  
  Values are chunk_summary, period_summary, conversation_summary, case_summary, case_context_filing_extraction, relevance_gate, evidentiary_map_element, or tag_suggestion.
  
  versionSlug string
  
  systemPrompt string
  
  tool object
  
  Hide tool attributes Show tool attributes object
  
  name string
  
  description string
  
  input_schema object
  
  model string
  
  maxTokens number
  
  temperature number
- statusCode number Required
- timestamp string(date-time) Required

GET /ai/admin/prompts/{purpose}/active

curl \
 --request GET 'http://api.example.com/ai/admin/prompts/{purpose}/active' \
 --header "Authorization: Bearer $ACCESS_TOKEN"

Response examples (200)

{
  "data": {
    "purpose": "case_summary",
    "versionSlug": "case_summary.v2",
    "systemPrompt": "string",
    "tool": {
      "name": "string",
      "description": "string",
      "input_schema": {}
    },
    "model": "sonnet",
    "maxTokens": 1024,
    "temperature": 0.2
  },
  "statusCode": 200,
  "timestamp": "2026-01-15T10:15:00.000Z"
}