memorious-mcp

Memorious MCP Server

memorious-mcp gives AI assistants persistent local memory for preferences, facts, and project information. Its store, recall, and forget tools use short keys and ChromaDB similarity search to manage information across sessions through MCP.

Server facts

memorious-mcp

Project

License
Not specified

Tech

Runtime / language
Not specified
Transports
stdio
Remote available
No
Authentication
Not specified

Activity

GitHub stars
9
Last commit
Data as of

Tags

  • Python
  • chromadb
  • embedding
  • key value
  • memory
  • semantic

Overview

Personal preferences and project facts can remain available between assistant sessions with memorious-mcp, a local memory service exposed through MCP. ChromaDB supplies persistent disk storage and vector-based retrieval, while FastMCP provides the tool implementation. The store tool pairs a compact canonical key with a value containing the full information and any context. Recall searches for related memories rather than relying on identical wording, returning stored values that can inform later responses. This supports workflows such as retaining user preferences, preserving conversational context, and retrieving project knowledge.

Key consistency shapes both retrieval and removal: queries should use the same concise, concept-focused style as stored keys. Recall accepts a top_k parameter to limit returned matches and produces an empty results collection when nothing matches. Forget uses a query key and top_k to select nearby memories for deletion, then returns the identifiers removed. Deletion cannot be reversed, so ambiguous requests call for confirmation. The service uses stdio for local client integrations, with storage and embedding processing handled on the user's machine.

memorious-mcp setup notes

  • memorious-mcp records stdio only; confirm a network transport upstream before planning remote access.
  • memorious-mcp has no recorded license; confirm reuse terms in its repository.

Related MCP servers