OMEM runs on your own machine — the client starts it, so there is no endpoint to ping. 138 installs a week from pypi. Last commit 14 Sep 2026.
Memory for AI agents that tracks beliefs over time and surfaces contradictions
What the code does
We read the source, 3 d ago · rules 995398eba69f
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.
этот файл ставится пользователю, но в репозитории его нет
self.__BUILD_MANIFEST={__rewrites:{afterFiles:[],beforeFiles:[],fallback:[]},"/_error":["static/chunks/pages/_error-7ce03bcf1df914ce.js"],sortedPages:["/_app","/_error"]},self.__BUILD_MANIFEST_CB&&self.__BUILD_MANIFEST_CB();
target = os.path.join(dest, os.path.relpath(root, source))
return subprocess.Popen([sys.executable, "-m", "omem.mcp_server"],
…e[o]}}),Qe("at");var Ze=function(t,e){return b(i[t].prototype[e])},tr=(Ze("Array","at"),TypeError),er=function(t,e){if(!delete t[e])throw new tr("Cannot delete property "+Y(e)+" of "+Y(t))},rr=Math.min,nr=[].copyWithin||function(t,e){var r=it(this),n=de(r),o=he(t,n),i=he(e,n),a=a…
Is this your server and something here is wrong? Tell us — corrections are free and do not require a plan.
This code can reach further than it looks
We found places where it runs commands, builds paths or queries from values it is given. None of that is a flaw by itself — it becomes one when the code changes, and code changes quietly between releases. We re-read it on every one.
Connect this server
This server runs on your own machine — install it with the package manager and the client starts it for you. Package name taken from the official registry entry.
claude mcp add omem -e 'OMEM_API_KEY=<OMEM_API_KEY>' -e 'OMEM_PROJECT=<OMEM_PROJECT>' -- uvx omem-infrastructure{
"mcpServers": {
"omem": {
"command": "uvx",
"args": [
"omem-infrastructure"
],
"env": {
"OMEM_API_KEY": "<OMEM_API_KEY>",
"OMEM_PROJECT": "<OMEM_PROJECT>"
}
}
}
}[mcp_servers.omem]
command = "uvx"
args = ["omem-infrastructure"]
env = { OMEM_API_KEY = "<OMEM_API_KEY>", OMEM_PROJECT = "<OMEM_PROJECT>" }{
"mcpServers": {
"omem": {
"command": "uvx",
"args": [
"omem-infrastructure"
],
"env": {
"OMEM_API_KEY": "<OMEM_API_KEY>",
"OMEM_PROJECT": "<OMEM_PROJECT>"
}
}
}
}{
"servers": {
"omem": {
"type": "stdio",
"command": "uvx",
"args": [
"omem-infrastructure"
],
"env": {
"OMEM_API_KEY": "<OMEM_API_KEY>",
"OMEM_PROJECT": "<OMEM_PROJECT>"
}
}
}
}Put your own values in place of
<OMEM_API_KEY> (API key from the OMEM server. It is printed on first run of `omem-server`), <OMEM_PROJECT> (Project id to read and write memory in, also printed on first run).
The author’s registry entry marks them as required.
Optional settings from the registry entry, left out of the block above:
OMEM_BASE_URL (URL of the OMEM server), OMEM_AGENT (The agent whose memory this process reads and writes. Fixed here rather than passed as a tool argument, so the model cannot act as a different agent), OMEM_USER (The end user this process is acting for. Also fixed here, not a tool argument. Leave unset and no user-scoped memory is visible).
Alternatives to OMEM
same job, measured the same wayA claims memory for agents: assert beliefs, let them decay, surface contradictions.
Agent-agnostic memory backend that preserves continuity between humans and AI over time.
Persistent memory for AI agents — log and recall conversation context over MCP.
Cognitive memory for AI agents — semantic recall, knowledge graph, and contradiction detection
Session memory for AI coding agents. Search past sessions, extract rules, and track what works.
MCP server that sanitizes bad broker data and compresses options chains 95% for AI trading agents.
Coordination server for parallel coding agents: claims, contracts, notices, and memory notes.
Tasks, comments, shared memory and handoffs for teams of people and AI agents, over MCP.
OMEM — questions
Answers built from our own checks of this server.