# TechDebtAssassin

- **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:** [chiranjeevi joshi](https://cerebralvalley.ai/u/Chiranjeevijoshi)
- **GitHub:** https://github.com/jchiru21/tech-debt-assassin
- **Demo video:** https://www.loom.com/share/f4ad1749a4e94fc48f70bd23413daead
- **Gallery:** https://cerebralvalley.ai/e/claude-code-hackathon/hackathon/gallery
- **Page:** https://cerebralvalley.ai/e/claude-code-hackathon/hackathon/gallery/70

TechDebtAssassin is an autonomous AI agent that eliminates the "janitorial" busywork of technical debt. Unlike standard coding assistants that wait for prompts, TechDebtAssassin is a proactive specialist that modernizes entire repositories without human intervention.
Powered by Claude Opus 4.6 and the Model Context Protocol (MCP), the agent builds a Global Project Context to reason across file boundaries—inferring complex type hints and architectural dependencies that local-context LLMs miss. It features a robust "Self-Healing" pipeline: scanning the AST for debt, generating type-safe patches, applying them physically to the disk, and running a final verification loop (using mypy) to ensure 100% code health.
Built for Problem Statement One, TechDebtAssassin transforms hours of manual refactoring into a one-click autonomous process, turning codebase maintenance from a chore into a reliable DevOps stage.

---

Markdown version of https://cerebralvalley.ai/e/claude-code-hackathon/hackathon/gallery/70. Site index for agents: https://cerebralvalley.ai/llms.txt · full text: https://cerebralvalley.ai/llms-full.txt