Sign in

v29

latestOpenAPI 3.1.0raw.githubusercontent.com2026-03-233765161.7 KB
v1 data

Ingest data for processing with memory integration

Ingest data for asynchronous processing

Accepts various content types (text, messages, files) and processes them to extract information
and integrate it into the user's memory system. Returns a job_id for tracking status.

**Entity Resolution:**
- user_id (str, required): Always required - the main user identifier
- persona_id (str, optional): If provided, data is ingested to this persona instead of user
- project_id (str, optional): If provided, data is ingested to this project (inherits from user)

Priority: persona_id > project_id > user_id

**Request Parameters:**
- content_type (str, required): One of: "text", "messages", "pdf", "word", "image", "video", "audio", "file"
- payload (str|dict|list, required): Content data (text string, message list, or base64 for files)
- content_description (str, optional): Description of the content being ingested (e.g., 'Logo design concepts', 'Meeting notes')
- session_id (str, optional): Groups related content for session-based retrieval
- timestamp (str, optional): ISO-8601 timestamp for historical data
- filename (str, optional): Original filename for file uploads

**Response:**
- job_id (str): Unique identifier for tracking the processing job
- user_id (str): Confirmed entity ID (user, persona, or project)
- content_type (str): Confirmed content type
- status (str): Job status ('queued', 'accepted')
- message (str): Status message
- created_at (str): ISO-8601 timestamp
- success (bool): True if accepted

**Example:**
```json
{
    "user_id": "user-123",
    "persona_id": null,
    "project_id": "project-456",
    "content_type": "text",
    "payload": "Meeting notes from today's discussion",
    "content_description": "Meeting notes from today's discussion"
}
```

Returns 202 Accepted with job_id. Use /job/status to check processing status.
Max payload: 5MB (JSON), 20MB (multipart). Requires JWT authentication.
post/v1/data/ingest

Request body

user_idstring required

User ID (always required)

persona_idstring nullable

Optional persona ID. If provided, data is ingested to this persona instead of the user

project_idstring nullable

Optional project ID. If provided, data is ingested to this project (inherits from user)

content_typestring nullable

Content category: 'text', 'image', 'video', 'pdf', 'audio', 'messages', 'file'. If omitted, the category is auto-detected from the uploaded file bytes.

content_descriptionstring nullable

Optional description of the content being ingested (e.g., 'Logo design concepts', 'Meeting notes')

session_idstring nullable

Session ID for grouping related ingested content and enabling session-based retrieval

timestampstring nullable

ISO-8601 timestamp to preserve original data moment

filenamestring nullable

Filename of the uploaded file

callback_urlstring nullable

Optional URL the server will POST to when the job reaches a terminal state (done, error, cancelled). The payload will match the /v1/data/job/status response shape.

notification_emailstring email nullable

Optional email address to notify when the job reaches a terminal state.

Example request

{
  "content_description": "Email correspondence about project updates",
  "content_type": "text",
  "payload": "From: john@example.com\nTo: jane@example.com\nSubject: Hello\n\nHello Jane!",
  "timestamp": "2024-01-01T12:00:00Z",
  "user_id": "user-123"
}

Response

Successful Response

job_idstring required

Unique job identifier for tracking

statusstring required

Processing status ('accepted', 'queued', 'failed')

messagestring nullable

Additional status or error message

successboolean

Whether the request was accepted successfully

Example response

{
  "content_type": "text",
  "created_at": "2024-01-01T12:00:00Z",
  "job_id": "123e4567-e89b-12d3-a456-426614174000",
  "message": "Content successfully uploaded and queued for processing",
  "status": "queued",
  "success": true,
  "user_id": "abc-123"
}
All 37 operations