Claude-Mem - Open Source Claude Code Memory Plugin with Cross-Session Persistent Memory Support

What's Claude-Mem?

Claude-Mem is for Claude Code designed as an open source plugin to address the problem of memory loss in AI across sessions. It helps Claude maintain continuity of knowledge about a project over a long period of time by automatically capturing the tool's use of observations, generating semantic summaries, and injecting relevant context in subsequent sessions. Core features include persistent memory, progressive recall, intelligent search tools, Web Viewer UI, and more. Users can install and use it with simple commands to automate operations without manual intervention.Claude-Mem supports privacy control and contextual configuration, allowing users to make fine-grained settings according to their needs. Claude-Mem provides an experimental Endless Mode, which significantly extends session length and reduces the number of context windows in the token Claude-Mem, through its unique compression and memorization features, provides a unique solution to the use of Claude Code of users provides a more efficient and continuous work experience.

Features of Claude-Mem

• Cross-session persistent memory: Automatically stores and manages key information from each session, including tool call results, conversation logs, project structure and development history, allowing the AI to re-explain the context in a new session without needing to do so.
• gradual recall: Presenting the memory in a three-layer structure of index → summary → original content, AI loads only the needed parts, saving token and avoiding information noise.
• Intelligent Search Tools: Provides skill-based searches, such as mem-search, that can query project history in natural language, saving about 2250 tokens at the start of each session.
• Web Viewer UI: Provides real-time memory stream visualization, available at http://localhost:37777查看.
• Privacy Controls: Use<private>Labeling excludes sensitive content storage.
• Context Configuration: Fine-grained control over the context of the injection.
• automatic operation: No manual intervention is required; the automated capture tool uses the observations and generates semantic summaries.
• reference function: This can be done byclaude-mem://URI references past decisions.
• Beta Channel: Try experimental features such as Endless Mode via version switching.

Claude-Mem's core strengths

• persistent memory: Save project context across sessions to avoid reinterpreting the context each time and improve efficiency.
• Intelligent Search and Cost Savings: Quickly query project history in natural language with the mem-search tool, saving about 2250 tokens per session.
• gradual recall: Layered retrieval of memorized content, on-demand loading, balancing information volume and cost.
• Automatic recording and operation: Automatically captures information about tool calls, file operations, etc., eliminating the need for manual archiving and reducing repetitive work.
• Privacy and Security: Support <private> Labeling to exclude sensitive content storage and ensure data privacy.
• Real-time visualization: Visualize the progress of your project by viewing the memory stream in real time through the Web Viewer UI.
• Flexible Configuration: Users can customize settings such as context injection to meet individual needs.
• experimental function: Offer Beta channels, such as Endless Mode, to extend session lengths and explore more possibilities.

What is Claude-Mem's official website

• Github repository:: https://github.com/thedotmack/claude-mem

People for Claude-Mem

• software developer: Developers who need to continually interact with AI in long-term projects can keep their code development consistent with persistent memory features, reducing the time spent repeating explanations of context.
• Technical Team: Teams that need to share project knowledge and decisions in multi-person collaboration, Claude-Mem can help team members quickly understand the progress of the project and improve collaboration efficiency.
• research worker: Researchers who frequently use AI-assisted analysis in complex research projects save time and maintain continuity of research ideas by quickly reviewing research history with intelligent search tools.
• Product Team: Teams responsible for product development and iteration can use Claude-Mem to document key decisions and changes made during product development for subsequent auditing and optimization.
• Data Analyst: People who need to track and review the analysis process over time in their data analysis projects, Claude-Mem helps them to quickly locate historical analysis results and increase their productivity.
• AI users: Anyone who uses Claude Code for complex tasks, especially those who need to maintain context across sessions, can significantly improve their experience with Claude-Mem.