Hey Product Hunt! I'm José, and I built Recall because I was tired of re-explaining my codebase to Claude Code every morning. The problem: every Claude Code session starts from zero. Your AI forgets your architecture, your decisions, your patterns. If you hit the 200K context limit, compaction summarises everything — and your nuanced decisions become one-liners. The solution: four Auto-Memory Hooks that capture context silently and reload it on every session start. Now available as a native Claude Code plugin: /plugin install recall@claude-plugins-official /recall:setup ← connects your API key in 30 seconds What makes this different from other MCP memory servers: 1. Native plugin — install like any Claude Code extension, auto-updates 2. Auto-hooks — memory capture is fully automatic, not manual 3. Survives compaction — the pre-compact hook saves state before context summarisation 4. Webhooks — route GitHub/Stripe events directly into Claude 5. Team sharing — one Claude learns, all Claudes know 6. Self-hosted — Docker + Redis, $12/mo, your data stays with you Pricing: Free (500 memories) / Pro $5/mo (7-day trial, no card) / Team $15/mo / Self-Hosted $12/mo. Open source core: github.com/joseairosa/recall I'd love your feedback — especially on the hook installation experience and what features would make this indispensable for your workflow.

recallmcp.com

Persistent memory for CC - install once, remember forever

Persistent memory for CC - install once, remember forever