{
  "$schema": "https://themachinepress.com/schemas/story-v1.schema.json",
  "schema_version": "1.0.0",
  "document_type": "machine_press_story",
  "story": {
    "story_id": "mp-2026-10-04-005",
    "source_story_id": "tmp-story-finskillbench-tools-documents",
    "edition_id": "mp-2026-10-04-morning-0087",
    "edition_url": "https://themachinepress.com/edition/2026-10-04",
    "position": 5,
    "story_type": "dispatch",
    "section": "business-enterprise",
    "editorial_classification": "editorial",
    "headline": "The Skill Package’s Biggest Advantage Was Its Calculator",
    "slug": "the-skill-package-s-biggest-advantage-was-its-calculator",
    "dek": "FinSkillBench separates procedural documents from executable tools in verifiable financial-agent tasks.",
    "summary": "FinSkillBench separates procedural documents from executable tools in verifiable financial-agent tasks.",
    "body_text": "FinSkillBench contains 2,603 dated episodes across twelve subtasks with deterministic verifiers. A paired analysis across eight models found curated skill packages raised mean scores by 16.2 points, while skills generated within one episode added only half a point and used more tokens and turns. Testing resources separately, documents added 5.6 points and executable tools 19.5; combining them was subadditive. A second harness reproduced the direction while changing effect sizes. The authors frame the premium as a property of the full model, resources and harness, rather than proof that the underlying model acquired financial expertise.",
    "why_it_matters": "FinSkillBench separates procedural documents from executable tools in verifiable financial-agent tasks.",
    "limitations": [
      "A paired analysis across eight models found curated skill packages raised mean scores by 16.2 points, while skills generated within one episode added only half a point and used more tokens and turns."
    ],
    "importance": 8,
    "canonical_url": "https://themachinepress.com/story/mp-2026-10-04-005/the-skill-package-s-biggest-advantage-was-its-calculator",
    "json_url": "https://themachinepress.com/story/mp-2026-10-04-005.json",
    "first_published_at": "2026-10-04T09:00:00.000-04:00",
    "modified_at": "2026-10-05T08:46:40.372-04:00",
    "content_status": "updated",
    "is_carryover": false,
    "carryover_reason": null,
    "key_claims": [
      {
        "claim_id": "claim-mp-2026-10-04-005-001",
        "text": "FinSkillBench separates procedural documents from executable tools in verifiable financial-agent tasks.",
        "source_ids": [
          "source-2026-10-04-005"
        ],
        "qualification": "A paired analysis across eight models found curated skill packages raised mean scores by 16.2 points, while skills generated within one episode added only half a point and used more tokens and turns."
      }
    ],
    "source_ids": [
      "source-2026-10-04-005"
    ],
    "tags": [
      "agent skills",
      "executable tools",
      "verifiable workflows"
    ],
    "image_url": null,
    "corrections": [
      {
        "correction_id": "correction-2026-10-04-001",
        "revision": 2,
        "summary": "Publication note: This October 4 edition was reconstructed on October 5 after the workstation was offline. It uses original October 2 submissions from the standard Sunday-evening arXiv release, rather than claiming those papers were available for the missed 9 AM run. The date and 9 AM publication slot preserve archive chronology; October 5 developments are excluded. The September 27 Invention Desk is carried over because its replacement was generated October 5.",
        "published_at": "2026-10-05T08:46:40.372-04:00",
        "affected_story_ids": [],
        "affected_claim_ids": []
      }
    ]
  },
  "sources": [
    {
      "source_id": "source-2026-10-04-005",
      "title": "arXiv preprint 2610.03564, original v1",
      "publisher": "arXiv / submitting research authors",
      "url": "https://arxiv.org/abs/2610.03564v1",
      "canonical_url": "https://arxiv.org/abs/2610.03564v1",
      "source_type": "primary_research",
      "is_primary_source": true,
      "published_at": "2026-10-02T12:44:04.000-04:00",
      "accessed_at": "2026-10-05T08:40:15.113-04:00",
      "supports_claim_ids": [
        "claim-mp-2026-10-04-005-001"
      ]
    }
  ],
  "corrections": [
    {
      "correction_id": "correction-2026-10-04-001",
      "revision": 2,
      "summary": "Publication note: This October 4 edition was reconstructed on October 5 after the workstation was offline. It uses original October 2 submissions from the standard Sunday-evening arXiv release, rather than claiming those papers were available for the missed 9 AM run. The date and 9 AM publication slot preserve archive chronology; October 5 developments are excluded. The September 27 Invention Desk is carried over because its replacement was generated October 5.",
      "published_at": "2026-10-05T08:46:40.372-04:00",
      "affected_story_ids": [],
      "affected_claim_ids": []
    }
  ],
  "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 Package’s Biggest Advantage Was Its Calculator",
    "publisher": "The Machine Press",
    "published_at": "2026-10-04T09:00:00.000-04:00",
    "canonical_url": "https://themachinepress.com/story/mp-2026-10-04-005/the-skill-package-s-biggest-advantage-was-its-calculator"
  }
}