Memphora
Attach Memphora’s memory layer so your assistant remembers user-specific facts and project notes across separate chat threads.
Overview
Memphora is a MCP server for the Build phase that lets AI assistants store and recall information across conversations via Memphora’s memory API.
What is this MCP server?
- PyPI package memphora-mcp v0.1.3 with python -m memphora_mcp.server
- Store and recall information across conversations for AI assistants
- MEMPHORA_API_KEY (required) plus optional MEMPHORA_USER_ID for memory namespacing
- stdio transport for standard MCP clients
- Version 0.1.3; registry PyPI identifier memphora-mcp; 2 documented env vars (API key required, user id optional).
Community signal: 3 GitHub stars.
What problem does it solve?
Research and scope notes live only in old chats, so your assistant cannot reliably continue multi-day solo-builder work.
Who is it for?
Solo builders using Python-friendly MCP setups who want a dashboard-backed memory service with a simple API key flow.
Skip if: Builders who need only in-repo CLAUDE.md context with zero external memory SaaS.
What do I get? / Deliverables
Facts you store through Memphora MCP are available in later sessions, keeping validation and build threads aligned.
- Cross-session store and recall for assistant-visible memory
- Optional per-user memory isolation via MEMPHORA_USER_ID
- MCP bridge to Memphora without custom API glue in each project
Recommended MCP Servers
Journey fit
Memphora is cataloged under Build because MCP registration is part of extending your agent stack during product work. agent-tooling covers third-party memory services that plug into Claude Code–style workflows via stdio MCP.
How it compares
Hosted assistant-memory MCP, not an in-editor rules file or single-shot prompt template.
Common Questions / FAQ
Who is Memphora for?
Developers and founders using MCP clients who want Memphora to persist and recall notes across many assistant conversations.
When should I use Memphora?
Use it when you split idea research, scoping, and coding across sessions and need the agent to remember what you already captured.
How do I add Memphora to my agent?
Add the PyPI memphora-mcp server with module memphora_mcp.server, set MEMPHORA_API_KEY, and optionally MEMPHORA_USER_ID in your MCP environment.