Claude memory: the top 10 ways to give Claude persistent, long-term memory, compared. Covers MCP memory servers, CLAUDE.md, memory plugins and tools for Claude Code, Claude Desktop and the Claude API. Setup, pros and cons, and which to pick for remembering context across sessions.
MCPpedia last refreshed this data
Claude Memory is an MCP server that claude memory: the top 10 ways to give Claude persistent, long-term memory, compared. Covers MCP memory servers, CLAUDE.md, memory plugins and tools for Claude Code, Claude Desktop and the Claude API. Setup, pros and cons, and which to pick for remembering context across sessions. Its tool list has not been published yet over stdio, sse and http, requires no API key, and scores 48/100 on MCPpedia's security, maintenance and efficiency rubric.
Config is the same across clients - only the file and path differ.
{
"mcpServers": {
"claude-memory": {
"args": [
"mem0ai"
],
"command": "uvx"
}
}
}Are you the author?
Add this badge to your README to show your security score and help users find safe servers.
Claude memory, compared. This guide covers every practical way to give Claude, Claude Code and Claude Desktop persistent, long-term memory across sessions: memory MCP servers, Claude Code plugins, CLAUDE.md, and Anthropic's built-in memory. Each entry has what it stores, where the memory lives, how to install it, and who it's for.
This server supports HTTP transport. Be the first to test it - help the community know if it works.
Five weighted categories - click any category to see the underlying evidence.
Mem0: Cross-tenant conversation leak via unescaped delimiter collision in session-scope key (_build_session_scope)
Summary _build_session_scope() in mem0/memory/main.py builds the conversation-history scope key used to fetch/store the last N raw messages by naively joining caller-supplied user_id/agent_id/run_id values with unescaped & and = delimiters: def _build_session_scope(filters): parts = [] for key in sorted(["user_id", "agent_id", "run_id"]): val = filters.get(key) if val: parts.append(f"{key}={val}") return "&".join(parts) _validate_and_trim_entity_id() (t
mem0ai Databricks vector store vulnerable to SQL injection
mem0 0.1.117 before 1.0.8 contains an sql injection vulnerability in mem0/vector_stores/databricks.py that allows attackers to inject SQL by influencing stored memory content. Attackers can supply crafted memory text, metadata, or message names that reach unescaped insert() and update() string interpolation to execute injected SQL on the configured Databricks SQL warehouse.
mem0 server lacks authentication and authorization controls for its memory creation API endpoint
The mem0 1.0.0 server lacks authentication and authorization controls for its memory creation API endpoint (POST /memories). The endpoint allows unauthenticated users to submit arbitrary memory records without verifying their identity or permissions. A remote attacker can exploit this by sending unauthenticated POST requests to create malicious or spoofed memory entries in the database, leading to unauthorized data injection and potential data pollution.
>= 0source →mem0 server lacks authentication and authorization controls for its memory deletion API endpoint
The mem0 1.0.0 server lacks authentication and authorization controls for its memory deletion API endpoint (DELETE /memories). The endpoint allows unauthenticated users to delete memory records by specifying arbitrary user identifiers (e.g., user_id, run_id, agent_id) in the request query parameters. A remote attacker can exploit this by sending unauthenticated DELETE requests to erase memory data for any user, leading to unauthorized data loss and denial of service.
>= 0source →mem0 server lacks authentication and authorization controls for its memory management API endpoints
The mem0 1.0.0 server lacks authentication and authorization controls for its memory management API endpoints. Critical functions such as updating memory records (PUT /memories/{memory_id}) are exposed without any verification of the requester's identity or permissions. A remote attacker can exploit this by sending unauthenticated requests to modify, overwrite, or delete arbitrary memory records, leading to unauthorized data manipulation and potential data loss.
>= 0source →Be the first to review
Have you used this server?
Share your experience - it helps other developers decide.
Sign in to write a review.
Others in ai-ml
🛡️ BLA (Blind LLM Agents) — Adversarial defense shield that protects binaries, Electron apps, JS bundles, Python, and WebAssembly against AI reverse-engineering agents like REA. Poison -> Exhaust -> Hallucinate.
Unified real-time search engine skill for AI agents.
An open-source AI agent that brings the power of Gemini directly into your terminal.
Codebase knowledge graph for AI agents — 162 languages, sub-ms queries, 99% fewer tokens.
MCP Security Weekly
Get CVE alerts and security updates for Claude Memory and similar servers.
Start a conversation
Ask a question, share a tip, or report an issue.
Sign in to join the discussion.
Claude memory, compared. This guide covers every practical way to give Claude, Claude Code and Claude Desktop persistent, long-term memory across sessions: memory MCP servers, Claude Code plugins, CLAUDE.md, and Anthropic's built-in memory. Each entry has what it stores, where the memory lives, how to install it, and who it's for.
Last updated: November 2026.
| # | Claude memory tool | Type | Memory is shared with | Where memory lives | Best for |
|---|---|---|---|---|---|
| ⭐ 1 | Glen | Plugin + CLI + remote MCP server | Your whole team | Hosted | Teams running Claude Code, Codex and Cursor |
| 2 | Claude Code CLAUDE.md + auto memory | Built in | Project (via git) / you | Local files | Every Claude Code user |
| 3 | Claude app memory | Built in | You | Anthropic | Chatting in claude.ai and Claude Desktop |
| 4 | claude-mem | Claude Code plugin | You | Local (~/.claude-mem) | Solo developers in Claude Code |
| 5 | Mem0 | Memory API + CLI | Your agents | Hosted or self-hosted | Building memory into your own app |
| 6 | Supermemory | Remote MCP server + API | You or a team space | Hosted or local | One memory across many AI clients |
| 7 | Basic Memory | MCP server | You | Local Markdown files | Notes you can read and edit by hand |
| 8 | MCP Knowledge Graph Memory | MCP server | You | Local JSON file | A minimal, self-hosted start |
| 9 | Cognee | Claude Code plugin + MCP server | You or your agents | Local or Cognee Cloud | Graph memory over your own documents |
| 10 | Graphiti / Zep | MCP server | Your agents | Self-hosted graph DB / Zep cloud | Facts that change over time |
[!TIP] Most Claude memory tools remember for one person. Glen remembers for the whole team. Every Claude Code session your teammates run is recorded, turned into memory, and fed back into your Claude while it works. If someone already figured out why the billing worker batches at 500, your Claude knows before you ask.
npx @tryglen/cli@latest install
Glen is shared memory for coding agents. It records the agent sessions on your team (Claude Code, Codex, Cursor, Pi and your own agents) and the tools you connect, such as Slack, GitHub, tickets, docs and meeting recorders (Fireflies, Fathom, Granola). From that it builds observations about people, projects and decisions, and puts the relevant ones into Claude's context on every turn.
Most Claude memory tools stop at "Claude remembers what I told it last week." Glen covers what the person at the next des