MCP Tools¶
All tools are available when engram is connected as an MCP server (engram-mcp).
Tool Reference¶
| Tool | Description |
|---|---|
engram_remember |
Store memory with type, priority, tags, namespace |
engram_recall |
Search episodic memories (compact format by default) |
engram_think |
Reason across episodic + semantic memory via LLM |
engram_status |
Show memory statistics |
engram_get_memory |
Retrieve full memory content by ID or prefix |
engram_timeline |
Chronological context around a memory |
engram_add_entity |
Add entity node to knowledge graph |
engram_add_relation |
Add relationship edge between entities |
engram_query_graph |
Query knowledge graph |
engram_ingest |
Dual ingest: extract entities + store memories |
engram_meeting_ledger |
Record structured meeting (decisions, action items) |
engram_feedback |
Record positive/negative feedback on memories |
engram_auto_feedback |
Auto-detect feedback from conversation context |
engram_cleanup |
Delete all expired memories |
engram_cleanup_dedup |
Deduplicate similar memories by cosine similarity |
engram_summarize |
Summarize recent N memories via LLM |
engram_session_start |
Begin new conversation session |
engram_session_end |
End active session with optional summary |
engram_session_summary |
Get summary of completed session |
engram_session_context |
Retrieve memories from active session |
engram_ask |
Smart query — auto-routes to recall or think |
Tool Details¶
engram_remember¶
Store a memory with optional metadata.
Parameters:
| Parameter | Type | Description |
|---|---|---|
content |
string | The memory content |
memory_type |
string | fact, decision, preference, todo, error, context, workflow, meeting_ledger |
priority |
int | 1-10, default 5 |
tags |
list[str] | Optional tags for filtering |
namespace |
string | Tenant namespace |
expires |
string | Expiry: 2h, 1d, 7d |
topic_key |
string | Unique key for upsert (replaces existing) |
engram_recall¶
Search episodic memories by semantic similarity.
Parameters:
| Parameter | Type | Description |
|---|---|---|
query |
string | Search query |
limit |
int | Max results, default 5 |
memory_type |
string | Optional type filter |
tags |
list[str] | Optional tag filter |
resolve_entities |
bool | Resolve pronoun/entity references |
resolve_temporal |
bool | Resolve date/time expressions |
engram_think¶
Run LLM reasoning across all memory (episodic + semantic + federated).
Parameters:
| Parameter | Type | Description |
|---|---|---|
question |
string | The question to reason about |
engram_ingest¶
Extract entities from messages and store as memories. This is the primary tool for bulk ingestion — it runs entity extraction first and only stores messages that contain meaningful entities (entity-gated).
Parameters:
| Parameter | Type | Description |
|---|---|---|
messages |
list | List of {role, content} message objects |
engram_meeting_ledger¶
Record a structured meeting with decisions and action items.
Parameters:
| Parameter | Type | Description |
|---|---|---|
title |
string | Meeting title |
decisions |
list[str] | Decisions made |
action_items |
list[str] | Tasks to follow up |
attendees |
list[str] | Participants |
topics |
list[str] | Topics discussed |
engram_feedback¶
Record feedback to adjust memory confidence scores.
- Positive: +0.15 confidence
- Negative: -0.2 confidence
- 3x negative + low confidence: auto-delete
Parameters:
| Parameter | Type | Description |
|---|---|---|
memory_id |
string | Memory ID |
feedback |
string | positive or negative |
Session Tools¶
Session tools track context within a conversation:
Sessions allow the agent to retrieve only the memories from the current conversation, separate from the full memory store.