Claude Memory Manager
Give Claude Code persistent memory across chats so solo builders do not re-explain repos, decisions, and preferences every session.
Overview
Claude Memory Manager is an MCP server for the Operate phase that stores and semantically recalls Claude Code context across sessions.
What is this MCP server?
- Captures context from Claude Code sessions for later reuse
- Semantic search over stored memories instead of scrolling old threads
- Explicit recall tools for pulling prior decisions and project facts
- Stdio MCP package via PyPI (claude-memory-manager 0.1.1, uvx)
- Targets Claude Code specifically—not a generic notes app
- Server version 0.1.1 on PyPI identifier claude-memory-manager
- Transport type stdio with runtimeHint uvx
- Repository github.com/NyxToolsDev/claude-memory-manager
Community signal: 1 GitHub stars.
What problem does it solve?
Every new Claude Code session starts cold, so indie builders waste time re-explaining the same project context and lose thread-local decisions.
Who is it for?
Solo developers who live in Claude Code daily and want durable, searchable memory without building a custom vector store.
Skip if: Teams that already standardize everything in docs and issue trackers and rarely depend on chat-local context.
What do I get? / Deliverables
After you register the server, Claude can search and pull prior session context on demand so iteration stays continuous instead of repetitive.
- Searchable memory store tied to your Claude Code workflow
- Recalled context injected into new sessions on demand
- Reduced re-onboarding time when returning to a stale repo
Recommended MCP Servers
Journey fit
Memory sits after you ship and keep building—production iteration depends on retained context, not one-off prompts. Iterate is where you refine workflows and agent setup; cross-session recall is an operational habit for daily agent use.
How it compares
Persistent agent memory MCP server, not a project documentation skill or generic embedding pipeline you host yourself.
Common Questions / FAQ
Who is Claude Memory Manager for?
It is for indie builders and power users of Claude Code who need cross-session recall of project context, decisions, and working notes.
When should I use Claude Memory Manager?
Use it when you switch tasks or sessions often and want semantic search over what you already told the agent instead of re-prompting from scratch.
How do I add Claude Memory Manager to my agent?
Add the stdio MCP entry for claude-memory-manager from PyPI (runtimeHint uvx) in your Claude Code MCP config per the server registry schema 0.1.1.