Memories API
Overview
Access and query extracted atomic facts (memories) from team communications.
GET /api/memories
List all memories/facts with filtering and pagination.
Query Parameters
Prop
Type
Response
{
"memories": [
{
"id": "550e8400-e29b-41d4-a716-446655440000",
"memory_text": "The team decided to use JWT tokens for authentication.",
"quality_score": 0.85,
"tier": "atomic",
"channel_id": "C12345",
"author_name": "Jane Smith",
"message_ts": "2026-04-13T10:30:00Z",
"topic_tags": ["authentication", "security"],
"entity_tags": ["JWT", "OAuth"],
"importance": "high",
"fact_type": "decision",
"source_message_id": "1234567890.123456",
"permalink": "https://slack.com/archives/..."
}
],
"total": 856,
"page": 1,
"pages": 43
}GET /api/memories/{memory_id}
Get a specific memory by ID.
Parameters
Prop
Type
Response
{
"id": "550e8400-e29b-41d4-a716-446655440000",
"memory_text": "The team decided to use JWT tokens for authentication.",
"quality_score": 0.85,
"tier": "atomic",
"cluster_id": "topic_123",
"channel_id": "C12345",
"platform": "slack",
"author_id": "U12345",
"author_name": "Jane Smith",
"message_ts": "2026-04-13T10:30:00Z",
"thread_ts": null,
"source_message_id": "1234567890.123456",
"topic_tags": ["authentication", "security"],
"entity_tags": ["JWT", "OAuth"],
"action_tags": [],
"importance": "high",
"graph_entity_ids": ["entity_jwt_decision"],
"fact_type": "decision",
"thread_context_summary": "After discussing trade-offs...",
"valid_at": "2026-04-13T10:30:00Z",
"invalid_at": null,
"superseded_by": null,
"potential_contradiction": false,
"source_lang": "en"
}POST /api/memories/search
Advanced search across memories with filters.
Request Body
{
"channel_id": "C12345",
"query": "authentication security",
"filters": {
"importance": ["high", "medium"],
"fact_types": ["decision"],
"entities": ["JWT"],
"date_range": {
"start": "2026-01-01T00:00:00Z",
"end": "2026-04-13T23:59:59Z"
}
},
"sort": "relevance",
"limit": 20
}Response
{
"results": [
{
"memory": {...},
"score": 0.95,
"highlights": ["The team decided to use <mark>JWT</mark> tokens..."]
}
],
"total": 15,
"query_time_ms": 45
}GET /api/memories/topics
List all topic tags across memories.
Parameters
Prop
Type
Response
{
"topics": [
{"tag": "authentication", "count": 45},
{"tag": "deployment", "count": 32},
{"tag": "api", "count": 28}
]
}GET /api/memories/entities
List all entity tags across memories.
Response
{
"entities": [
{"name": "JWT", "count": 23},
{"name": "PostgreSQL", "count": 18},
{"name": "Redis", "count": 15}
]
}Memory Quality Scores
Memories are scored 0.0 to 1.0 based on:
- Information Density: Specific vs. general content
- Actionability: Contains decisions or action items
- Specificity: Concrete details vs. vague statements
- Recency: Adjusted for staleness over time
Quality Bands
| Score | Quality | Description |
|---|---|---|
| 0.8-1.0 | High | Decisions, concrete action items, specific technical details |
| 0.5-0.8 | Medium | Opinions, observations, contextual information |
| 0.0-0.5 | Low | General statements, ambiguous content |
Memory Types
| Type | Description | Examples |
|---|---|---|
decision | Decisions made, conclusions reached | "We chose PostgreSQL over MongoDB" |
opinion | Opinions, preferences, suggestions | "I think we should use TypeScript" |
observation | Observations, facts noted | "The API response time increased" |
action_item | Tasks to be done | "Need to update the documentation" |
question | Questions asked | "What's the token expiration time?" |
How is this guide?