Agent Memory runs on your own machine — the client starts it, so there is no endpoint to ping. Last commit 22 Sep 2026.
Persistent agent memory with a decision layer: replay, restore, verify, or none — explainable.
Today is the operative word: we check Agent Memory every 15 minutes and re-read its code on every release. Watch it and you find out the day that stops being true.
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 agent-memory -- uvx agent-memory-sdk
{
"mcpServers": {
"agent-memory": {
"args": [
"agent-memory-sdk"
],
"command": "uvx"
}
}
}
[mcp_servers.agent-memory]
command = "uvx"
args = ["agent-memory-sdk"]
{
"mcpServers": {
"agent-memory": {
"args": [
"agent-memory-sdk"
],
"command": "uvx"
}
}
}
{
"mcpServers": {
"agent-memory": {
"args": [
"agent-memory-sdk"
],
"command": "uvx"
}
}
}
This one needs environment variables set before it will start:
AGENT_MEMORY_DIR (Directory for the persistent memory store (default: ~/.agent_memory)), AGENT_MEMORY_COLLECTION (Collection (database) name within the memory directory).
The author declared them in the registry entry; get the values from the project itself.
Local-first persistent memory for AI agents, with reasoning and formal verification.
Persistent project memory for LLMs via SQLite. Never re-explain your project again.
Persistent memory for AI agents. Search and store durable facts, preferences and decisions.
Open-source causal memory for AI agents: persistent, explainable, MCP-native memory (13 tools).
Persistent memory for AI agents. Memories decay with time and strengthen on recall.
Verifiable agent memory with signed ledger receipts — verify every write offline.
Evidence-backed, explainable persistent memory for MCP-compatible AI agents.
Persistent cross-platform memory for AI agents with Ebbinghaus decay scoring.
Answers built from our own checks of this server.