{
  "id": 13687,
  "url": "https://arxiv.org/abs/2607.22520v1",
  "title": "The Regression Tax: Decomposing Why Skills Help and Hurt LLM Agents",
  "summary": "Adding procedural skills to an LLM agent is typically evaluated by average improvement in task success. However, this metric hides an important cost: skills can also make agents worse. We measure both sides by comparing agents with and without skills across nearly 6,000 runs spanning two office automation benchmarks and three model harness stacks. This allows us to distinguish two outcomes. A regression is a task solved without skills but failed after skills are added. A residual failure is a ta",
  "authors": "Darshan Tank, Baran Nama",
  "category": "research",
  "topics": "jobs-economy,agents-autonomy",
  "orgs": null,
  "regions": null,
  "published_at": "2026-07-24T17:50:03.000Z",
  "fetched_at": "2026-07-27T05:10:06.638Z",
  "source_slug": "x-arxiv-cs-ai",
  "source_name": "arXiv cs.AI",
  "source_homepage": "https://arxiv.org/list/cs.AI/recent",
  "ethics_ai_record_url": "https://ethics.ai/record/13687",
  "original_url": "https://arxiv.org/abs/2607.22520v1",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}