Enterprise DNA
M MCP Servers Developer low

Patdolitse/piia-engram

by Various

One memory. Every AI tool. Yours to keep. Local-first, MCP-compatible, Apache 2.0.

P

MCP

Patdolitse/piia-engram

Added 1 June 2026

#ai-agent #ai-identity #ai-identity-layer #ai-memory #anthropic #claude-code #claude-desktop #codex

Overview

Piia-Engram is a local-first memory layer for AI tools that stores and retrieves context across sessions using an MCP-compatible interface. It runs as a Python service, keeping user data on their own machine under the Apache 2.0 license.

Best for

Best for
Developers who want a private, self-hosted memory layer for MCP-compatible AI tools

Use cases

  • Persist conversation history across different AI chat interfaces
  • Share context between multiple MCP-compatible tools without cloud storage
  • Build a private, self-hosted memory backend for AI agents

Notes

Piia-Engram is a local-first memory layer for AI tools that stores and retrieves context across sessions using an MCP-compatible interface. It runs as a Python service, keeping user data on their own machine under the Apache 2.0 license.

161 stars on GitHub. Last updated 2026-06-01. Licensed Apache-2.0.

Use cases

  • Persist conversation history across different AI chat interfaces
  • Share context between multiple MCP-compatible tools without cloud storage
  • Build a private, self-hosted memory backend for AI agents

Pros

  • Local-first design keeps sensitive data on your own machine
  • MCP compatibility allows integration with many existing AI tools
  • Open source with permissive Apache 2.0 license

Cons

  • Requires Python runtime and manual setup to run
  • Limited to MCP-compatible tools, not a universal memory solution
  • Small community (161 stars) means fewer examples and support

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.

Pros

  • Local-first design keeps sensitive data on your own machine
  • MCP compatibility allows integration with many existing AI tools
  • Open source with permissive Apache 2.0 license

Cons

  • Requires Python runtime and manual setup to run
  • Limited to MCP-compatible tools, not a universal memory solution
  • Small community (161 stars) means fewer examples and support