MCP Knowledge Graph
mcp-knowledge-graph
Persistent memory for AI in a local knowledge graph: entities, relations and observations are stored in .aim files on your disk
- 888 ★ on the repository
- Last commit
- MIT
Extended access
Why this level
- The server reads and writes memory files on disk
- The files may hold personal notes and work details
Install
- Works in
Manual install
Connect mcp-knowledge-graph per the README, create an .aim folder in the project root and check that memory is written to .aim/memory.jsonl. Add .aim to .gitignore if it holds personal data.
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The server reads and writes memory files on disk
- The files may hold personal notes and work details
This is third-party code. Review the repository files before installing.
What it does
MCP Knowledge Graph keeps model memory between conversations as a graph: entities, relations and observations. Data lives in JSONL files. If the project root has a `.aim` folder, `.aim/memory.jsonl` from the project is used, otherwise a file in the global folder set by the `--memory-path` flag. Contexts can be separated: `memory-work.jsonl`, `memory-personal.jsonl`. It works with Claude Code, Claude Desktop and other MCP clients. Version 1.3.2 from December 2025, MIT license.
Who it is for. For people who want an agent to remember a project and preferences between sessions, with the memory kept in readable files on their disk.
Good fit when
- You need agent memory between sessions in local files
- You need separate memory per project through the .aim folder
- You need separate contexts: work and personal
Not a fit when
- You need cloud memory reachable from different machines: here the files are on disk
- You need semantic search: here is a graph of entities and relations
Example request
Remember that in this project tests run with pnpm test and we name branches feature/<task>MCP
- Transport
- stdio
- Authentication
- not required
Limitations
Memory is written to JSONL files on your disk, and they may contain everything the agent decided to remember: names, work details, personal notes. Do not commit an .aim folder with personal data to a shared repository. The folder name .aim matters for project detection. Access is extended: the server reads and writes memory files.
How to disable. Remove the server entry from the client settings and the .aim or memory*.jsonl files if the memory is no longer needed.
FAQ
What does MCP Knowledge Graph store?
Entities, relations and observations in JSONL files: locally in the project .aim folder or in a global folder.
How is the memory file chosen?
If the project has .aim, .aim/memory.jsonl is used, otherwise the global folder from --memory-path.
Can contexts be separated?
Yes, files get a suffix: memory-work.jsonl, memory-personal.jsonl.
Discussion
No questions or reviews yet.
Related
MCP serversiai-memory
iai-memory (Personal Memory Engine)
A local long-term memory MCP server for agents: keeps conversations verbatim and returns the right context at the start of a session
Self-hosted long-term memory for agents: semantic search, a knowledge graph and autonomous consolidation over MCP and a REST API
Long-term agent memory in local Markdown files with a link graph, semantic search and MCP access
Local persistent memory for coding agents: hooks record every session, and Claude Code, Codex or Cursor recall past decisions through 54 MCP tools
An MCP memory server for agents built on a knowledge graph: relevance as an evidence path, not just vector similarity
A long-term memory system for AI agents: the agent stores experience and facts, then retrieves them through a built-in MCP server
Comparison with similar tools
Entries that share the main search query, and how they differ.
| Entry | Official | Authentication | Repository stars | Main difference |
|---|---|---|---|---|
| MCP Knowledge Graph (this entry) | no | not required | 888 | |
| agent-memory | no | 3,404 | long-term memory on Markdown files with a SQLite index | |
| OpenWolf | no | 2,387 | shared local project memory with a project map and token tracking | |
| MemoraX Code | no | 2,192 | a memory plugin for architecture and repository rules, version 0.1.x |
README in short
The README describes persistent memory through a local knowledge graph with entities, relations and observations, storage in .aim and the --memory-path flag. MIT license.