Osiris runs on your own machine — the client starts it, so there is no endpoint to ping. Last commit 2 Nov 2026.
Cache-aware context engine for AI agents: server-side payload filtering, honest context ledger.
Today is the operative word: we check Osiris 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 osiris -- npx -y osiris-mcp
{
"mcpServers": {
"osiris": {
"command": "npx",
"args": [
"-y",
"osiris-mcp"
]
}
}
}
[mcp_servers.osiris]
command = "npx"
args = ["-y", "osiris-mcp"]
{
"mcpServers": {
"osiris": {
"command": "npx",
"args": [
"-y",
"osiris-mcp"
]
}
}
}
{
"servers": {
"osiris": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"osiris-mcp"
]
}
}
}
Optional settings from the registry entry, left out of the block above:
OSIRIS_HOME (Repertoire d'etat OSIRIS : base SQLite, rapports, caches. Defaut ~/.osiris), OSIRIS_ROOTS (Racines autorisees pour la lecture de fichiers, separees par ';'. Vide = aucune lecture de fichier), OSIRIS_CACHE_TTL_S (TTL du cache en secondes), OSIRIS_PRUNE_THRESHOLD_TOKENS (Seuil de prune en tokens), OSIRIS_DISPATCH_ONLY (1 = n'expose que l'outil osiris.call (table d'outils stable pour le prefix cache)).
Local-first persistent context memory MCP server for AI coding agents.
Local context and engineering evidence for AI coding agents.
Context-aware code retrieval MCP server — ranked, gated results for AI coding agents
MCP server for cache.overflow - AI agents sharing knowledge with AI agents
Temporal engineering context for agents via MCP.
On-demand, local code-context MCP server for AI coding agents with Tree-sitter AST.
Memory, docs, and repo context layer for engineering agents.
Context Assurance for AI agents: context engineering, AI cost optimization, verification.
Answers built from our own checks of this server.