MCP Light Memory runs on your own machine — the client starts it, so there is no endpoint to ping. Last commit 27 Aug 2026.
Local-first persistent memory for coding agents and MCP clients.
What the code does
We read the source, 12 h ago · rules 3dff92dd89df
A value the model can set ends up inside a file or shell call. That is not a flaw by itself — for a terminal server it is the job — but it is where things go wrong when it is not.
const p = Bun.spawn(cmd, { cwd: context.worktree, stdout: "pipe", stderr: "pipe" })
What this server is able to do. For an MCP server this is often the job itself — a terminal server runs commands because that is what it is for. Listed so you know what you are plugging in, not as an accusation.
out = subprocess.check_output(
p = Bun.spawn([py, script, ...args], {
Is this your server and something here is wrong? Tell us — corrections are free and do not require a plan.
A tool parameter here reaches a dangerous call
That is not a flaw by itself — but it is where things go wrong when it is not the job. We re-read this code on every release. Watch it and you hear from us the day another one appears.
Alternatives to MCP Light Memory
same job, measured the same wayLocal-first persistent memory for AI coding agents.
Persistent memory for AI coding agents. #1 on LongMemEval. Local-first.
Local-first, persistent memory for AI coding agents — remember, recall, forget, and list, over MCP.
Persistent memory MCP server for AI coding assistants. Chinese-first and fully local.
Persistent memory, coordination, and learning for AI agents. Local-first, 25 MCP tools.
Local-first persistent memory for AI agents — SQLite + local embeddings, MCP-native, no cloud.
Local-first, state-centric memory for coding agents. MCP server + CLI.
Local-first, cross-harness memory for AI coding agents.
MCP Light Memory — questions
Answers built from our own checks of this server.