Ossian MCP runs on your own machine — the client starts it, so there is no endpoint to ping. 39 installs a week from npm. Last commit 12 Oct 2026.
Ask your own documents with citations, plus durable memory for agents. Safe by default.
Today is the operative word: we check Ossian MCP 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 ossian-mcp -- npx -y ossian-mcp
{
"mcpServers": {
"ossian-mcp": {
"args": [
"-y",
"ossian-mcp"
],
"command": "npx"
}
}
}
[mcp_servers.ossian-mcp]
command = "npx"
args = ["-y", "ossian-mcp"]
{
"mcpServers": {
"ossian-mcp": {
"args": [
"-y",
"ossian-mcp"
],
"command": "npx"
}
}
}
{
"mcpServers": {
"ossian-mcp": {
"args": [
"-y",
"ossian-mcp"
],
"command": "npx"
}
}
}
This one needs environment variables set before it will start:
OSSIAN_URL (Base URL of the Ossian installation (e.g. http://localhost:8081).), OSSIAN_API_KEY (Ossian API key (osk_…). Scope it to the narrowest role/namespace that works.), OSSIAN_AGENT_ID (Separates one agent's memories from another's. Defaults to 'mcp'.), OSSIAN_MODE (Access mode: read-only | read-write | admin. Starts read-only; writes need read-write.).
The author declared them in the registry entry; get the values from the project itself.
Shared, source-backed memory for AI agents on your own Postgres, with an optional document engine.
Build a RAG pipeline station by station and search your own documents with cited answers.
Hybrid search over your local documents, with file and page citations. One binary, fully local.
Private memory, documents, voice and a local LLM for your agents, served by a Mac app on-device.
City council meetings, transcripts, votes, and documents for U.S. cities, with citations.
Omitly's local, verifiable PDF redaction for AI agents — your documents never leave your device.
Local-first memory for AI agents: recall with sources, answer checks, documents, mesh, governance.
Self-hosted memory for AI agents with hybrid search, knowledge graphs, tasks, and document sync.
Answers built from our own checks of this server.