# Juancho

- **Event:** [Built with Opus 4.6: a Claude Code hackathon](https://cerebralvalley.ai/e/claude-code-hackathon)
- **When:** Feb 10 at 12:00 PM – Feb 17 at 10:00 AM (EST)
- **Where:** Location TBA
- **Team:** [Juan Munoz Corredor](https://cerebralvalley.ai/u/jmuncor)
- **GitHub:** https://github.com/jmuncor/mumpu
- **Demo video:** https://www.youtube.com/watch?v=oUMxUSKiu-8
- **Gallery:** https://cerebralvalley.ai/e/claude-code-hackathon/hackathon/gallery
- **Page:** https://cerebralvalley.ai/e/claude-code-hackathon/hackathon/gallery/203

Every LLM conversation starts from zero: you re-explain who you are, what you're working on, and what you prefer, every single time. Mumpu is a transparent HTTP proxy that gives any LLM tool (Claude CLI, Cursor, ChatGPT, etc.) long-term memory with zero code changes. It extracts the key atomic facts from your conversations, deduplicates them intelligently, and on every future request, uses LLM-powered semantic search to retrieve matching memories and injects them into the system prompt automatically. Works with OpenAI, Anthropic, and Gemini through a single port, includes a real-time knowledge graph dashboard, and is grounded in 7 research papers. Built entirely with Claude Opus 4.6 via Claude Code.

---

Markdown version of https://cerebralvalley.ai/e/claude-code-hackathon/hackathon/gallery/203. Site index for agents: https://cerebralvalley.ai/llms.txt · full text: https://cerebralvalley.ai/llms-full.txt