openapi: 3.1.0
info:
  title: Persona Ingestion Query API
  version: 1.0.0
  description: 'Persona is an open-source (MIT) user-memory system for AI agents, built by Eder Labs. It ingests raw content (conversations, notes) for a user, extracts a Graph-Vector hybrid memory (Neo4j graph + vector embeddings), and answers natural-language RAG queries or returns structured insights over that memory. This specification is transcribed faithfully from the published FastAPI router source (server/routers/graph_api.py) and the docs/API.md reference of github.com/saxenauts/persona. Persona is self-hosted: the base URL below is the documented local default; deploy behind your own host and auth layer.'
  license:
    name: MIT
    url: https://github.com/saxenauts/persona/blob/main/LICENSE
  contact:
    name: Persona (Eder Labs)
    url: https://docs.buildpersona.ai/
    email: utkarsh@buildpersona.ai
  x-apievangelist:
    source: https://github.com/saxenauts/persona/blob/main/server/routers/graph_api.py
    docs: https://docs.buildpersona.ai/
    method: searched
servers:
- url: http://localhost:8000/api/v1
  description: Documented self-hosted local default (uvicorn / docker compose)
tags:
- name: Query
  description: Retrieving memory via RAG and structured insights
paths:
  /users/{user_id}/rag/query:
    post:
      operationId: rag_query
      summary: RAG query
      description: Runs a retrieval-augmented-generation query over the user's memory and returns a natural-language answer. Query max length is 1000 characters.
      tags:
      - Query
      parameters:
      - $ref: '#/components/parameters/UserId'
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/RAGQuery'
      responses:
        '200':
          description: RAG answer
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/RAGResponse'
        '400':
          description: Missing or too-long query
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/Error'
        '404':
          description: User not found
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/Error'
        '502':
          description: External service (LLM) error
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/Error'
        '503':
          description: Database (Neo4j) connection error
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/Error'
  /users/{user_id}/ask:
    post:
      operationId: ask_insights
      summary: Ask for structured insights
      description: Asks a question over the user's memory and returns a structured result, optionally shaped by a caller-provided output_schema.
      tags:
      - Query
      parameters:
      - $ref: '#/components/parameters/UserId'
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/AskRequest'
      responses:
        '200':
          description: Structured insight result
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/AskResponse'
        '400':
          description: Missing request body or query
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/Error'
        '404':
          description: User not found
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/Error'
        '502':
          description: External service (LLM) error
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/Error'
        '503':
          description: Database (Neo4j) connection error
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/Error'
components:
  schemas:
    AskRequest:
      type: object
      required:
      - query
      properties:
        query:
          type: string
          example: What are my preferences?
        output_schema:
          type: object
          description: Optional caller-provided shape for the structured result.
          additionalProperties: true
    RAGResponse:
      type: object
      properties:
        answer:
          type: string
          example: Based on your memories, you're working on the Q4 roadmap...
    RAGQuery:
      type: object
      required:
      - query
      properties:
        query:
          type: string
          maxLength: 1000
          example: What projects am I working on?
    AskResponse:
      type: object
      properties:
        result:
          type: object
          additionalProperties: true
          example:
            preferences:
            - remote work
            - morning meetings
            summary: User prefers flexible work arrangements
    Error:
      type: object
      description: FastAPI default error envelope.
      properties:
        detail:
          type: string
          example: User alice not found
  parameters:
    UserId:
      name: user_id
      in: path
      required: true
      description: The unique identifier for the user. Pattern ^[a-zA-Z0-9_-]+$.
      schema:
        type: string
        pattern: ^[a-zA-Z0-9_-]+$