recall

Stop wasting Claude Code tokens every time you resume.

13 followers

Stop wasting Claude Code tokens every time you resume.

13 followers

Visit website

AI Metrics and Evaluation

•

Prompt Engineering Tools

•

LLM Memory

Claude Code starts every session cold. Recall keeps a local log of your sessions and condenses it into a resume-ready summary entirely on your machine. No API key, no external model, nothing sent anywhere. It's built for people running Claude Code locally on a subscription: the only AI in the loop is Claude Code itself; the summarization is done by a classical Python summarizer. It is Free on your subscription, Saves your usage credits and Nothing leaves your machine.

Free

Launch tags:Software Engineering•Developer Tools•Artificial Intelligence

Launch Team / Built With

Render Sandboxes: a computer for every agentRun untrusted code safely. Idle sandboxes cost nothing

Promoted

buildpipe

Maker

📌

👋 Hey Product Hunt! I built Recall to fix the most annoying part of my day with Claude Code: every single session starts cold. You spend a great session deep in a problem — Claude knows the goal, the files, where you got stuck, what's left to do. Then you close the terminal. Next morning you open a fresh session and... it remembers nothing. So you re-explain the project. Again. Burning tokens (and your patience) on context you already paid for once. The existing fixes never quite did it for me: - CLAUDE.md is hand-written notes you have to maintain — it doesn't record what actually happened. - --resume replays the whole transcript — full fidelity, but token-heavy and stuck to one machine. - Most "AI memory" tools quietly pipe your code, paths, and sometimes secrets to a model endpoint to summarize them. Hard no for a lot of people. Recall takes a different approach: 🔒 100% local. No API key, no external model, no network calls. Nothing leaves your machine — ever. The summarization runs as classical Python (TF-IDF + TextRank), not an LLM call. 💸 Costs you zero tokens. Because the summary is built by a local algorithm, capturing and updating your memory spends no model tokens. And resuming from a compact ~1–2K token digest instead of re-explaining everything actually stretches your subscription further. 📝 Two plain files in your project. history.md logs every session automatically as you work; context.md is the condensed "where are we right now" — goal, files touched, commands run, next steps, where you left off. Both are diffable and shareable (commit them for team memory, or keep them personal). ⚡ Zero friction. No pip install, no local model to run, no key to configure. It works offline and starts the moment the plugin loads. It's free, MIT-licensed, and it's its own marketplace: /plugin marketplace add raiyanyahya/recall /plugin install recall@recall I'd genuinely love your feedback — especially on what you'd want a session summary to capture. I'm in the comments all day. 🙏

Report

5mo ago

Forum Threads

p/recall-21

•

5mo ago

How are you handling the "cold start" problem in Claude Code today ?

Hey Product Hunt I'm the maker of Recall (github.com/raiyanyahya/recall).

The itch that started it: Claude Code begins every session cold. I was re-explaining my project goal, where I left off, what I'd already tried at the start of almost every session, and it felt like a tax I kept paying. So I built Recall to fix that without the thing I didn't want: shipping my transcripts off to some endpoint.

View all