{
  "$schema": "https://themachinepress.com/schemas/story-v1.schema.json",
  "schema_version": "1.0.0",
  "document_type": "machine_press_story",
  "story": {
    "story_id": "mp-2026-08-24-009",
    "source_story_id": "tmp-story-aces-skill-lift",
    "edition_id": "mp-2026-08-24-morning-0046",
    "edition_url": "https://themachinepress.com/edition/2026-08-24",
    "position": 9,
    "story_type": "dispatch",
    "section": "developer-tools",
    "editorial_classification": "editorial",
    "headline": "The Skill Had to Beat the Same Agent Without It",
    "slug": "the-skill-had-to-beat-the-same-agent-without-it",
    "dek": "ACES measures capability packages through paired live trials under a fixed model, workspace, sandbox and scorer.",
    "summary": "ACES measures capability packages through paired live trials under a fixed model, workspace, sandbox and scorer.",
    "body_text": "Across 947 paired cases from 58 production skills and four harnesses, the preprint reports a mean composite Skill Lift of 0.2134 and positive lift in 72.8 percent of cases. Static scans and runtime outcomes were only weakly correlated, suggesting that document quality and actual agent benefit are complementary gates.",
    "why_it_matters": "ACES measures capability packages through paired live trials under a fixed model, workspace, sandbox and scorer.",
    "limitations": [
      "Static scans and runtime outcomes were only weakly correlated, suggesting that document quality and actual agent benefit are complementary gates."
    ],
    "importance": 8,
    "canonical_url": "https://themachinepress.com/story/mp-2026-08-24-009/the-skill-had-to-beat-the-same-agent-without-it",
    "json_url": "https://themachinepress.com/story/mp-2026-08-24-009.json",
    "first_published_at": "2026-08-24T09:00:00.000-04:00",
    "modified_at": "2026-08-24T09:00:00.000-04:00",
    "content_status": "new",
    "is_carryover": false,
    "carryover_reason": null,
    "key_claims": [
      {
        "claim_id": "claim-mp-2026-08-24-009-001",
        "text": "ACES measures capability packages through paired live trials under a fixed model, workspace, sandbox and scorer.",
        "source_ids": [
          "source-2026-08-24-009"
        ],
        "qualification": "Static scans and runtime outcomes were only weakly correlated, suggesting that document quality and actual agent benefit are complementary gates."
      }
    ],
    "source_ids": [
      "source-2026-08-24-009"
    ],
    "tags": [
      "agent skills",
      "continuous evaluation",
      "runtime evidence"
    ],
    "image_url": null,
    "corrections": []
  },
  "sources": [
    {
      "source_id": "source-2026-08-24-009",
      "title": "arXiv preprint 2608.20614",
      "publisher": "arXiv",
      "url": "https://arxiv.org/abs/2608.20614",
      "canonical_url": "https://arxiv.org/abs/2608.20614",
      "source_type": "primary_research",
      "is_primary_source": true,
      "published_at": "2026-08-20T19:26:14.000-04:00",
      "accessed_at": "2026-08-24T08:23:00.000-04:00",
      "supports_claim_ids": [
        "claim-mp-2026-08-24-009-001"
      ]
    }
  ],
  "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 Skill Had to Beat the Same Agent Without It",
    "publisher": "The Machine Press",
    "published_at": "2026-08-24T09:00:00.000-04:00",
    "canonical_url": "https://themachinepress.com/story/mp-2026-08-24-009/the-skill-had-to-beat-the-same-agent-without-it"
  }
}