Claude Memory
Give Claude Code persistent, searchable memory with hybrid retrieval, local embeddings, and MCP-backed context management across sessions.
Overview
claude-memory is a plugin marketplace entry for the Build phase that adds persistent hybrid-search memory and MCP context management to Claude Code.
What is this marketplace?
- Marketplace listing for Claude Memory: persistent intelligent memory with hybrid search
- Local embeddings via Ollama or LM Studio style stacks (per tag set)
- MCP plus Skills hybrid integration for Claude Code context management
- Knowledge graph and session-management oriented tags in catalog metadata
- MIT license, Node >=22, Claude Code >=1.0 compatibility in requirements
- 0 plugins enumerated in marketplace.json-style listing (standalone memory product)
- Stated requirements: Node >=22, Claude Code >=1.0
- Listing version 1.0.0, MIT license, free pricing type
What problem does it solve?
Agent sessions reset context, forcing you to re-upload decisions and redo discovery every time you return to a project.
Who is it for?
Solo builders running long Claude Code engagements who want local-friendly embeddings, MCP tools, and durable project memory.
Skip if: Quick one-off prompts with no reuse, or teams that forbid local embedding runtimes and MCP extensions.
What do I get? / Deliverables
You register Claude Memory so embeddings-backed retrieval and session management carry project knowledge forward into new Claude Code work.
- Registered Claude Memory marketplace integration (0 sub-plugins; product delivered via hybrid Skills/MCP stack)
- Queryable persistent memory store with hybrid search for project context
- Session-oriented context management hooks for ongoing agent work
Recommended Marketplaces
Journey fit
Memory infrastructure extends how you build with agents—canonical shelf is Build → agent-tooling because it plugs into Claude Code as a hybrid Skills/MCP memory layer. Listing centers on embeddings, hybrid search, knowledge graph, and session management for the coding agent, not app frontend or payments.
How it compares
MCP-backed persistent memory product listing, not a Ship-phase review pack or a single static skill file.
Common Questions / FAQ
Who is Claude Memory for?
It is for Claude Code users who need durable, searchable project memory across sessions using hybrid search, embeddings, and MCP integration.
When should I use Claude Memory?
Use it during Build and Operate when multi-day agent work, repeating repositories, or knowledge-heavy tasks outgrow ephemeral chat context.
How do I add Claude Memory to my agent?
Add the bpmforge/bpm-memory-mcp marketplace in Claude Code, satisfy Node >=22 and Claude Code >=1.0, then configure the Claude Memory MCP and embedding backend per repository docs.