SuperLocalMemory MCP
Local memory preserves decisions and facts across agent sessions, returning stored records with sources and corrections.
- 271 ★ on the repository
- Last commit
- AGPL-3.0
Extended access
Why this level
- The server retains work data across sessions and edits memory records; optional cloud features send selected context to an external service.
Install
- Works in
Manual install
Install SuperLocalMemory on this computer for me. Follow https://github.com/qualixar/superlocalmemory/blob/main/SETUP_FOR_AGENTS.md step by step and show me what each step printed. Ask me only one question, before turning on anything optional: whether I also want to find my pictures, screenshots and PDFs by describing them. Connect SuperLocalMemory to the agent I am using. When you finish, run slm doctor and tell me in plain words what works.
The first README route is an agent-guided setup following https://github.com/qualixar/superlocalmemory/blob/main/SETUP_FOR_AGENTS.md, ending with slm doctor results.
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The server retains work data across sessions and edits memory records; optional cloud features send selected context to an external service.
This is third-party code. Review the repository files before installing.
What it does
SuperLocalMemory stores records an agent can recall in a later session. Users save project decisions, rules and events, then search through meaning, keywords and filters. Results can stay concise or expand into sources and associated corrections. Several connected agents share memory with project and profile boundaries. A dashboard provides search, record relationships and storage health. An optional local-model module adds pictures and documents.
Who it is for. Teams using several coding agents that need to preserve project decisions for later sessions.
Good fit when
- A new agent must find a previously agreed architecture decision.
- After correcting a fact, the user needs a traceable link to the old record.
- While switching between Claude Code and Codex, a user needs the same project decisions.
- Notes need a source-backed answer with a bounded amount of returned text.
Not a fit when
- A short rules file is enough without searchable storage.
- Work-session content must not be retained in memory.
- You expect proof of a fact that was never stored.
Example request
Find why this project chose SQLite, show the original decision record and check whether a later correction replaced it.MCP
- Transport
- stdio, http
- Authentication
- not required
Limitations
Local-only behavior depends on enabled features: cloud models, remote processing and online answer checks send the relevant data outside the machine. No prebuilt installation is offered for Intel Macs or 32-bit Windows, so verify platform support before downloading. Pictures and documents require 16 GB of memory and additional model downloads. Clients receive the full memory tool profile by default; choose a smaller profile for a limited task and restart connected applications.
How to disable. Disconnect SuperLocalMemory from the agent and stop its server. Separately choose which stored records, backups and external connections to remove.
FAQ
What if memory has no answer?
The server reports a missing record or labels nearby matches separately. Inspect their sources before asking the agent to proceed.
How do I open the memory dashboard?
Run `slm dashboard`.
Why are results incomplete after startup?
Models may still be loading. The warm-up message asks you to retry after a few minutes.
How do I check the installation?
Run `slm doctor` and resolve reported issues before adding work data.
Does the online answer check enable itself?
No. Jev needs separate consent and sends the question with the top three memories.
Discussion
No questions or reviews yet.
Related
MCP serversLong-term memory for agents: a remote MCP server, plugins for Claude Code, Cursor and Codex, and an API for your own apps
Local persistent memory for coding agents: hooks record every session, and Claude Code, Codex or Cursor recall past decisions through 54 MCP tools
iai-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
Local project memory for agents: constraints, decisions and lessons live in an evidence-backed tree and return before a task.
Local memory for coding agents: sessions turn into a git-versioned Markdown wiki and come back through an MCP server.
Shared local memory for AI agents: Claude Code, Codex, Cursor and others find the agreements of past sessions in it and record new ones
README in short
SQLite is the canonical store. Temporal filters, exports and retention controls are available; these controls do not constitute compliance certification.