memora-mcp

trwilcoxson/memora-mcp
★ 0 stars Python AI/LLM Updated 7d ago
MCP server exposing Microsoft Memora as persistent semantic memory for coding agents (Claude Code, Codex, Cursor, Omnigent)
View on GitHub → 🔍 Audit Wallet Slippage →

Quick Install

Copy the config for your editor. Some servers may need additional setup — check the README.

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "memora-mcp": {
      "command": "uvx",
      "args": [
        "memora-mcp"
      ]
    }
  }
}

Or install with pip: pip install memora-mcp

README Excerpt

Persistent, automatic memory for a **meta-harness** — one memory shared across every agent it runs. Backed by [Microsoft Memora](https://github.com/microsoft/Memora)'s harmonic memory and wired into [Omnigent](https://github.com/omnigent-ai/omnigent) through standard hooks. Work through any agent; the durable facts you establish are captured on their own and surface in later sessions — in *any* agent — recalled by meaning, not keyword. It runs on the model plane you already have (your Claude/Cod

Tools (10)

MEMORA_CAPTUREMEMORA_EMBEDDINGMEMORA_LLM_MODELMEMORA_MAX_DISTILLS_PER_DAYMEMORA_MCP_HOMEMEMORA_MCP_USERMEMORA_MIN_TURNSMEMORA_PLANEMEMORA_REASONINGMEMORA_SESSION_SUMMARY