{
  "$schema": "https://themachinepress.com/schemas/story-v1.schema.json",
  "schema_version": "1.0.0",
  "document_type": "machine_press_story",
  "story": {
    "story_id": "mp-2026-10-17-013",
    "source_story_id": "tmp-story-os-helix-computer-use",
    "edition_id": "mp-2026-10-17-morning-0099",
    "edition_url": "https://themachinepress.com/edition/2026-10-17",
    "position": 13,
    "story_type": "dispatch",
    "section": "developer-tools",
    "editorial_classification": "editorial",
    "headline": "The Desktop Agent Stops Treating Code as a Bolt-On",
    "slug": "the-desktop-agent-stops-treating-code-as-a-bolt-on",
    "dek": "OS-Helix trains GUI actions and code execution in one action space after naive tool addition made the visual agent worse.",
    "summary": "OS-Helix trains GUI actions and code execution in one action space after naive tool addition made the visual agent worse.",
    "body_text": "Desktop work crosses visible interfaces, files, scripts and hidden system state. OS-Helix trains GUI actions and code execution together, using application-artifact composition to build executable Ubuntu tasks with solution-independent final-state checks. Step-balanced reinforcement learning aggregates importance by action step so long code blocks do not dominate optimization. On OSWorld-Verified, the authors report 33.7 to 35 percent relative gains over matched hybrid baselines; the 27-billion-parameter RL version reached 66.44 percent success and transferred to WindowsAgentArena. These are benchmark results in constructed environments, not proof of safe unattended desktop operation.",
    "why_it_matters": "OS-Helix trains GUI actions and code execution in one action space after naive tool addition made the visual agent worse.",
    "limitations": [
      "Step-balanced reinforcement learning aggregates importance by action step so long code blocks do not dominate optimization.",
      "These are benchmark results in constructed environments, not proof of safe unattended desktop operation."
    ],
    "importance": 8,
    "canonical_url": "https://themachinepress.com/story/mp-2026-10-17-013/the-desktop-agent-stops-treating-code-as-a-bolt-on",
    "json_url": "https://themachinepress.com/story/mp-2026-10-17-013.json",
    "first_published_at": "2026-10-17T09:00:00.000-04:00",
    "modified_at": "2026-10-17T09:00:00.000-04:00",
    "content_status": "new",
    "is_carryover": false,
    "carryover_reason": null,
    "key_claims": [
      {
        "claim_id": "claim-mp-2026-10-17-013-001",
        "text": "OS-Helix trains GUI actions and code execution in one action space after naive tool addition made the visual agent worse.",
        "source_ids": [
          "source-2026-10-17-013"
        ],
        "qualification": "Step-balanced reinforcement learning aggregates importance by action step so long code blocks do not dominate optimization."
      }
    ],
    "source_ids": [
      "source-2026-10-17-013"
    ],
    "tags": [
      "computer use",
      "GUI agents",
      "code execution"
    ],
    "image_url": "https://themachinepress.com/issues/2026-10-17/legal-motion-planning.webp",
    "corrections": []
  },
  "sources": [
    {
      "source_id": "source-2026-10-17-013",
      "title": "arXiv preprint 2610.22234, original v1",
      "publisher": "arXiv",
      "url": "https://arxiv.org/abs/2610.22234v1",
      "canonical_url": "https://arxiv.org/abs/2610.22234v1",
      "source_type": "primary_research",
      "is_primary_source": true,
      "published_at": "2026-10-15T13:14:02.000-04:00",
      "accessed_at": "2026-10-17T08:26:35.645-04:00",
      "supports_claim_ids": [
        "claim-mp-2026-10-17-013-001"
      ]
    }
  ],
  "corrections": [],
  "publisher": {
    "name": "The Machine Press",
    "url": "https://themachinepress.com",
    "description": "A daily newspaper for the age of artificial intelligence."
  },
  "cite_this_report": {
    "title": "The Desktop Agent Stops Treating Code as a Bolt-On",
    "publisher": "The Machine Press",
    "published_at": "2026-10-17T09:00:00.000-04:00",
    "canonical_url": "https://themachinepress.com/story/mp-2026-10-17-013/the-desktop-agent-stops-treating-code-as-a-bolt-on"
  }
}