The latest articles on DEV Community by Quy Dinh (@dvquy13). https://dev.to/dvquy13 https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F3870744%2Ff606ed75-3dfa-493d-8adb-8f03971c0f79.png https://dev.to/dvquy13 en Quy Dinh Fri, 10 Apr 2026 01:44:36 +0000 https://dev.to/dvquy13/helping-claude-code-remember-how-i-built-a-local-session-recall-tool-4p11 https://dev.to/dvquy13/helping-claude-code-remember-how-i-built-a-local-session-recall-tool-4p11 <p>Every Claude Code session starts fresh. That's fine at first, but 1,000+ sessions in, I had weeks of decisions technically on disk and practically unreachable. The sessions live in <code>~/.claude/projects/</code> as raw JSONL packed with tool calls and thinking blocks—readable in theory, but nobody is trawling through that.</p> <p>I built <a href="https://github.com/dvquy13/qrec" rel="noopener noreferrer">qrec</a>: a local session recall tool that strips the noise, indexes everything with hybrid search (BM25 + semantic + RRF), and lets a Claude Code agent retrieve past decisions on demand. No tokens spent; your data stays on your machine.</p> <p>The use case I didn't design for ended up being the most useful: context handoff. Deep in a 200-turn session about to hit the context limit, I opened a fresh one and typed "pick up context from the previous session." It came back with the exact decision we'd been circling, and we continued at ~15% context usage.</p> <p>The full post covers what I tried first (QMD, claude-mem), the bets I made on transparency and local-only execution, and—honestly—why the eval is still unsolved.</p> <p><strong><a href="https://dvquys.com/posts/introducing-qrec/?utm_source=devto&amp;utm_medium=crosspost&amp;utm_campaign=introducing-qrec" rel="noopener noreferrer">Read it on my blog →</a></strong></p> claude devtools opensource