Jevmem adds automatic project memory to Claude Code
Jevmem is an open-source tool that automatically saves decisions, constraints, bugs, and todos from Claude Code chats into a structured markdown file called JEVMEM.md. Built on Jev, a system by TypeSafe AI, it integrates as a plugin for Claude Code, Cursor, and Codex, marking superseded lines instead of deleting them. The tool uses a lightweight LLM to decide whether to save a turn, with a…
Key points
- Jevmem saves Claude Code chats as structured markdown with timestamps, confidence scores, and superseded-line tracking
- Uses a fast, low-cost LLM (Jev) for decision-making, outperforming others in speed but trailing in accuracy
- Includes security features like data scrubbing and poisoning gate, but relies on TypeSafe’s terms for retention
Jevmem includes security features like scrubbing sensitive data (API keys, passwords) and a poisoning gate to block malicious instructions. It supports retries for Jev outages and offers CLI commands for initialization, auditing, and manual memory management. The project is early-stage, with benchmarks limited to held-out turns and no long-term recall testing. Users must provide a TypeSafe API key (OpenAI or Anthropic keys are optional).
Model pages: Jev → · GPT-6 (Astra) → · Claude Opus 5.5 →
Jevmem – automatic project memory for Claude Code, built on Jev
github.com · 22 September 2026Loading the full article…
This text was published by github.com. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
Coverage and discussion
1source- Hacker News discussion · 44 pointsnews.ycombinator.com
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Agents & Tools
All →- Google says Gemini autonomously accessed three companies, then stopped itself · 3 src
- Meta’s Muse leaks internal files via prompt injection, developers say · 2 src
- Simon Willison warns coding agents may complicate software engineering · 1 src
- Perplexity trains agent on real mistakes with Hint-Guided Self-Distillation · 1 src
- Apple Intelligence for Home costs up to three times more than Ring and Nest in tests · 1 src
Comments
via GitHub Discussions