Agentic Harness Engineering (AHE)
What it is
AHE is a Python framework for automatically evolving coding-agent harnesses by optimizing prompts, tools, and middleware based on observability data. It uses an 'Evolve Agent' to propose evidence-backed changes to the harness components after analyzing agent traces, aiming to improve task success rates without retraining the underlying LLM.
The problem it solves
Coding agents often underperform on complex tasks due to suboptimal harness configurations (prompts, tools, context handling); manual tuning is slow, non-scalable, and doesn't generalize well across models.
What you could use it for
- Improving the performance of coding agents on benchmarks like Terminal-Bench 2
- Automating the refinement of agent system prompts and tool definitions
- Transferring optimized agent configurations across different base models
Read the source ↗Back to the library →
Summary written by an AI model from the source; check the source before you rely on it.