Remembra Relay

remembra-ai/remembra

This is a self-hosted memory layer that gives Claude persistent recall across conversations through 11 MCP tools. It automatically extracts entities from what you tell it, resolves that "Sarah" and "Sarah from Acme" are the same person, and builds a temporal knowledge graph you can query later. You get hybrid semantic and keyword search, temporal decay patterns, and conflict detection when memories contradict each other. The quickstart script spins up Remembra with Qdrant and Ollama locally in one command, no API keys required. It includes a dashboard for browsing entities and relationships, plus Python and TypeScript SDKs if you want to integrate it into your own tools. Useful when you need Claude to remember user preferences, past decisions, or relationship context beyond the current chat window.

GitHub stars: 16
tools: 5
auth: Secret needed
transport: STDIO
env vars: 3
last updated: Nov 11, 2026

View on GitHub

Tools

5 tools

Public tool metadata for what this MCP can expose to an agent.

store_memoryStore information in persistent memory. Automatically extracts entities (people, organizations, locations) and facts from the content.3 params

Store information in persistent memory. Automatically extracts entities (people, organizations, locations) and facts from the content.

Parameters* required

contentstring

The text content to memorize. Can be facts, context, decisions, meeting notes, preferences, etc.

metadataobject

Optional key-value metadata to attach (e.g., {"source": "meeting", "date": "2024-01-15"})

ttlstring

Optional time-to-live. Examples: "24h", "7d", "30d", "1y", or omit for permanent storage.

recall_memoriesSearch persistent memory for relevant information using hybrid search (semantic + keyword). Use this BEFORE answering questions about past decisions, context, people, or projects.3 params

Search persistent memory for relevant information using hybrid search (semantic + keyword). Use this BEFORE answering questions about past decisions, context, people, or projects.

Parameters* required

limitinteger

Maximum number of memories to return (1-50, default: 5)default: 5

querystring

Natural language query. Can be a question or keywords.

thresholdnumber

Minimum relevance score 0.0-1.0 (default: 0.4)default: 0.4

forget_memoriesDelete memories from persistent storage. GDPR-compliant deletion. Provide exactly one of: memory_id, entity, or all_memories.3 params

Delete memories from persistent storage. GDPR-compliant deletion. Provide exactly one of: memory_id, entity, or all_memories.

Parameters* required

all_memoriesboolean

Set to true to delete ALL memories. Use with extreme caution!

entitystring

Delete all memories about a specific entity (person, org, etc)

memory_idstring

Delete a specific memory by its ID

health_checkCheck Remembra server health and connection status. Returns server version, status, and component health.

Check Remembra server health and connection status. Returns server version, status, and component health.

No parameter schema in public metadata yet.

ingest_conversationIngest a conversation and automatically extract memories. Processes messages and intelligently extracts facts worth remembering long-term.5 params

Ingest a conversation and automatically extract memories. Processes messages and intelligently extracts facts worth remembering long-term.

Parameters* required

extract_fromstring

Which messages to extract from: 'user', 'assistant', or 'both'one of user · assistant · bothdefault: both

messagesarray

List of message dicts with 'role' and 'content'. Optional: 'name', 'timestamp'.

min_importancenumber

Minimum importance threshold (0.0-1.0, default: 0.5)default: 0.5

session_idstring

Optional session ID for grouping related conversations

storeboolean

If false, returns extraction results without storing (dry run)default: true

Configuration

REMEMBRA_API_KEY*secret

Remembra API key (dashboard: Settings > API keys)

REMEMBRA_URL*default: https://api.remembra.dev

Remembra server URL: https://api.remembra.dev, or your self-hosted server

REMEMBRA_AGENT_ID

Name this agent's handoffs are recorded under, e.g. claude-code, codex, cursor

Readme

Remembra

Remembra Relay

One agent stops. The next one already knows.
When a connected agent's session ends, Remembra Relay saves a handoff (a session that did nothing leaves none).
With the session hooks it is built from git and, for Claude Code and Codex, the session transcript.
An agent that closes through MCP reports its own facts, marked unverified.
The next agent, on another machine, from another vendor, or next week, starts from it.

View the full README on GitHub

Details

Repository: github.com/remembra-ai/remembra
Website: remembra.dev
Category: AI & LLM Tools
Package: remembra-mcp
Transport: STDIO
Auth: Required
Registry: MCP registry, active

Related AI & LLM Tools MCP servers

View all

Elefanteamenophis1er/elefanteOpen, Git-native memory protocol for MCP agents. Portable, versioned, human-editable.15stars

Figbridgerudraptpsingh/figbridgeFree, local MCP bridge for Figma — HTML, CSS, Tailwind, tokens, agent handoff bundles.15stars

Openfuse Mcpopenfused/openfusedEncrypted mail, shared memory, and persistent context for AI agents.14stars

Context Memjubakitiashvili/context-memPersistent memory for AI agents — 98%+ retrieval recall, 99% token savings, 44 tools13stars

Synapse Layer — Trust Infrastructure for AI Agentssynapselayer/synapse-layerMCP-native Trust Infrastructure for AI Agents. Persistent encrypted memory with Trust Quotient.13stars

Hashnethashgraph-online/hashnet-mcp-jsUniversal AI Agent Discovery - search 59,000+ agents across MCP, NANDA, A2A, OpenRouter0stars