Back to AI News
arxiv

Agency: An Interposition Layer Cuts Agent Harness Costs by Up to 35%

The paper analyzes cost accumulation patterns across agent traces and argues that harness-side choices, while having minimal impact on task accuracy, strongly affect execution cost, with prior work reporting reductions up to 2.6x on the same model and task. It introduces Agency, a harness interposition platform offering a unified abstraction for portable optimizations on off-the-shelf harnesses such as Claude Code and OpenAI Codex. Proof-of-concept techniques built on this analysis reduce cost by up to 35% across real-world benchmarks without compromising accuracy or modifying the model.

AI Brief

It offers a practical systems approach to lowering LLM agent execution costs without sacrificing task accuracy, relevant to teams running agents in production.

arXiv:2611.12122v1 Announce Type: new Abstract: AI agent harnesses play the critical role of providing LLMs with an interface to a computing environment. Recent studies show that AI agents' task-execution accuracy depends mostly on the LLM used, while harness-side variances such as system prompts or tools have minimal impact. Rather, such harness-side variances show a larger impact on the cost of agent executions, achieving cost reductions up to 2.6 times on the same model and task. To further explore these harness-side cost optimization opportunities, we analyze the cost accumulation patterns across various agent traces and identify system opportunities to reduce token use and LLM calls without modifying the model or affecting task accuracy. Based on this analysis, we design Agency, a harness interposition platform that provides a unified abstraction and allows easy and portable implementation of harness-side system optimizations on existing off-the-shelf harnesses such as Claude Code or OpenAI Codex. Using Agency, we implement proof-of-concept optimization techniques based on our analysis and achieve cost reductions up to 35% across various real-world benchmarks on existing harnesses without compromising accuracy.
Be the first to rate this article.

Comments & Questions (0)

Captcha
Please be respectful — let's keep the conversation friendly.

No comments yet

Be the first to comment!