Memlord
Self-hosted MCP memory server for personal use and teams
Ads
Overview
- 🔍 Hybrid search — BM25 (full-text) + vector KNN (pgvector) fused via Reciprocal Rank Fusion - 📂 Multi-user — each user sees only their own memories; workspaces for shared team knowledge - 🛠️ 11 MCP tools — store, retrieve, recall, list, search by tag, get, update, delete, move, list workspaces, dream report - 💤 Dreaming — a guided consolidation pass (dream MCP prompt + dream_report tool): finds near-duplicate and conflicting memories, merges them into insights non-destructively, driven by the client LLM - 🌐 Web UI — browse, search, edit and delete memories in the browser; export/import JSON - 🔒 OAuth 2.1 — full in-process authorization server, always enabled - 🐘 PostgreSQL — pgvector for embeddings, tsvector for full-text search - 📊 Progressive disclosure — search returns compact snippets by default; call get_memory(name) only for what you need, reducing token usage - 🔁 Deduplication — automatically detects near-identical memories before saving, preventing noise accumulation
Installation
docker compose up
Configuration
Configuration example is not available yet.
Capabilities
- Store a memory with type and tags
Use `store_memory` to save a fact, preference, instruction, feedback, decision, or insight, optionally setting an expiration.
- Search memories with hybrid retrieval
Ask `retrieve_memory` to find relevant memories using combined BM25 full-text and vector semantic search, returning compact snippets.
- Recall memories by time expression
Use `recall_memory` to find memories matching natural-language time queries like "last week" or "yesterday."
- Browse and filter all memories
Call `list_memories` with type and tag filters to paginate through stored memories.
- Consolidate memories via dreaming
Run the `dream` prompt to have the assistant review `dream_report` candidates, merge near-duplicates into insights, and retire superseded memories non-destructively.
Tags
- memory
- knowledge-retrieval
- database
- collaboration
Related MCP servers
InstantDB
OfficialCreate, manage, and update applications on InstantDB, the modern Firebase.
Jitsu
OfficialManage Jitsu data pipelines from your AI assistant — destinations, streams, connections, syncs, functions, and live events.
PostgreSQL MCP Server
OfficialRun read-only SQL queries and inspect PostgreSQL schemas from your MCP client.
Financial Datasets
OfficialStock market API made for AI agents