Show HN: Agent Tuning, using recursion to achieve predictable agent output

Category: ai-ml

Tags: ai-agents, probabilistic-programming, agent-tuning

Score: 6.0/10 (Innovation: 7, Technical: 5, Documentation: 8, Utility: 4)

Agent Tuning explores the concept of treating Claude Code's .claude/ directories as probabilistic programs, using recursion and self-reflection to achieve predictable agent behavior. It's interesting because it provides a formal framework for understanding and tuning non-deterministic AI agents by analyzing attention mechanisms and context drift.

Target audience: ai-engineers, ml-researchers, backend-devs

Repository: https://github.com/adam-s/agent-tuning · Python · 6 stars

View on Hacker News