What problem it solves

Agent sessions often end without preserving the observations and project knowledge needed for continuity. Claude-Mem addresses this by capturing session activity, generating semantic summaries, and making relevant context available in later sessions.

Key features

• Automatic session capture — Claude-Mem records tool usage and other session observations through lifecycle hooks and a local worker service. The captured activity is stored without requiring manual intervention.

  Source: README, plugin

• Semantic memory and summaries — The system compresses captured observations into semantic summaries and keeps sessions, observations, and summaries in a persistent SQLite database. This provides a structured history of project activity.

  Source: README

• Progressive memory search — Memory can be queried through a layered workflow that starts with compact search results, adds timeline context, and retrieves full observations only for selected IDs. A Chroma vector database supports semantic and keyword retrieval.

  Source: README

• Configurable context injection — Users can control the context supplied to new sessions through settings in `~/.claude-mem/settings.json`. The system also supports workflow modes and multiple languages for generated observations and summaries.

  Source: README, plugin/modes

• Broad harness integration — The installer supports several agent environments, including Claude Code, OpenCode, T3 Code, Grok Bot, and other documented integrations. Installation can also be performed through the Claude Code plugin marketplace.

  Source: README, openclaw

Quick start

Copied verbatim from the project's README.

npx claude-mem install

// Step 1: Search for index
search(query="authentication bug", type="bugfix", limit=10)

// Step 2: Review index, identify relevant IDs (e.g., #123, #456)

// Step 3: Fetch full details
get_observations(ids=[123, 456])

Who it may suit

Inference: This project may suit developers and teams using Claude Code or other supported agent harnesses who want project context to remain available across sessions.

Topics to discuss

• Which contexts should be captured, summarized, and injected automatically into new sessions?

• How should privacy controls and private tags be handled across local and hosted providers?

• What changes would make the progressive search workflow more useful for project history?

Written by ReadmeX AI from the project's README, source and public GitHub data; it may contain mistakes.

Read the original README https://github.com/thedotmack/claude-mem#readme

ReadmeX opened this community for a trending project; it is not run by the project. Maintainers can claim it from the community page.