API for managing memory, context, and retrieval for AI agents or apps with authentication and user-specific data.
API for managing memory, context, and retrieval for AI agents or apps with authentication and user-specific data.
This API supports three authentication methods:
X-API-Key headerX-API-Key: <your-api-key>X-Session-Token headerX-Session-Token: <your-session-token>Authorization headerAuthorization: Bearer <token>All endpoints require one of these authentication methods.
Submit multiple feedback items in a single request.
Useful for submitting session-end feedback or bulk feedback collection. Each feedback item is processed independently, so partial success is possible.
Authentication Required: One of the following authentication methods must be used:
Authorization headerX-API-Key headerX-Session-Token headerRequired Headers:
List of feedback items to submit
The search_id from SearchResponse that this feedback relates to
The feedback data containing all feedback information
A pointer to a Parse object
A pointer to a Parse object
Types of feedback that can be provided
Where the feedback was provided from
Internal user ID (if not provided, will be resolved from authentication)
Internal user ID (if not provided, will be resolved from authentication)
Internal user ID (if not provided, will be resolved from authentication)
External user ID for developer API keys acting on behalf of end users
External user ID for developer API keys acting on behalf of end users
External user ID for developer API keys acting on behalf of end users
Optional organization ID for multi-tenant feedback scoping. When provided, feedback is scoped to this organization.
Optional organization ID for multi-tenant feedback scoping. When provided, feedback is scoped to this organization.
Optional organization ID for multi-tenant feedback scoping. When provided, feedback is scoped to this organization.
Optional namespace ID for multi-tenant feedback scoping. When provided, feedback is scoped to this namespace.
Optional namespace ID for multi-tenant feedback scoping. When provided, feedback is scoped to this namespace.
Optional namespace ID for multi-tenant feedback scoping. When provided, feedback is scoped to this namespace.
curl -i -X POST \
http://memory.papr.ai/v1/feedback/batch \
-H 'Content-Type: application/json' \
-H 'X-API-Key: YOUR_API_KEY_HERE' \
-d '{
"feedback_items": [
{
"external_user_id": "dev_api_key_123",
"feedbackData": {
"assistantMessage": {
"__type": "Pointer",
"className": "PostMessage",
"objectId": "abc123def456"
},
"citedMemoryIds": [
"mem_123",
"mem_456"
],
"citedNodeIds": [
"node_123",
"node_456"
],
"feedbackImpact": "positive",
"feedbackProcessed": true,
"feedbackScore": 1,
"feedbackSource": "inline",
"feedbackText": "This answer was very helpful and accurate",
"feedbackType": "thumbs_up",
"feedbackValue": "helpful",
"userMessage": {
"__type": "Pointer",
"className": "PostMessage",
"objectId": "abc123def456"
}
},
"search_id": "abc123def456",
"user_id": "abc123def456"
}
],
"session_context": {}
}'{ "code": 200, "status": "success", "feedback_ids": [ "fb_123", "fb_456" ], "successful_count": 2, "failed_count": 0, "errors": [], "message": "Processed 2 feedback items successfully" }
Retrieve feedback by ID.
This endpoint allows developers to fetch feedback details by feedback ID. Only the user who created the feedback or users with appropriate permissions can access it.
Authentication Required: One of the following authentication methods must be used:
Authorization headerX-API-Key headerX-Session-Token headerRequired Headers:
curl -i -X GET \
'http://memory.papr.ai/v1/feedback/{feedback_id}' \
-H 'X-API-Key: YOUR_API_KEY_HERE'{ "code": 200, "status": "success", "feedback_id": "fb_123456789", "message": "Feedback retrieved successfully", "details": { "feedback_type": "thumbs_up", "feedback_score": 1, "feedback_text": "This was helpful!", "search_id": "search_123", "created_at": "2024-01-17T17:30:45.123456Z" } }
Upload and process documents using the pluggable architecture.
Authentication Required: Bearer token or API key
Supported Providers: TensorLake.ai, Reducto AI, Gemini Vision (fallback)
Features:
Preferred provider for document processing.
Preferred provider for document processing.
curl -i -X POST \
http://memory.papr.ai/v1/document \
-H 'Authorization: Bearer <YOUR_JWT_HERE>' \
-H 'Content-Type: multipart/form-data' \
-F file=string \
-F preferred_provider=gemini \
-F hierarchical_enabled=true \
-F schema_id=string \
-F simple_schema_mode=false \
-F graph_override=string \
-F property_overrides=string \
-F namespace=string \
-F user_id=string \
-F end_user_id=string \
-F webhook_url=string \
-F webhook_secret=string \
-F metadata=stringDocument upload started
Human-readable status message
Human-readable status message
Human-readable status message
Error message if failed
Error message if failed
Error message if failed
Additional error details or context
Additional error details or context
Additional error details or context
Status and progress of the document upload
Post ID in Parse Server (user-facing page ID)
Post ID in Parse Server (user-facing page ID)
Post ID in Parse Server (user-facing page ID)
Processing status type
Processing status type
Processing status type
{ "code": 200, "status": "success", "message": "string", "error": "string", "details": {}, "document_status": { "progress": 0, "current_page": 0, "total_pages": 0, "current_filename": "string", "upload_id": "string", "page_id": "string", "status_type": "processing", "error": "string" }, "memory_items": [ { … } ], "memories": [ { … } ] }