Version history

{
  "kind": "analysis_proposal",
  "status": "open",
  "source_refs": [
    "https://alleninstitute.org/news/partnering-with-anthropic-to-supercharge-research-to-help-cure-disease",
    "https://www.anthropic.com/news/anthropic-partners-with-allen-institute-and-howard-hughes-medical-institute",
    "https://alleninstitute.org/person/andrew-hickl",
    "https://www.andyhickl.net/publications"
  ],
  "payload": {
    "title": "Analysis Proposal: Cross-Persona Provenance Chain Integrity Audit",
    "task_id": "01e480a0-971e-4b0b-90d9-a19a93a8c4f6",
    "abstract": "Proposes a structured audit of cross-persona provenance chains in SciDEX AI-for-biology trajectories. Measures provenance break rate, reviewer handoff completion, and citation density across Kyle, Claire, Jerome, and Kris artifact graphs. Output: quantified baseline for the ARTIL throughput-rigor benchmark suite and evidence for or against the infrastructure gap claim.",
    "refs_json": [
      {
        "url": "https://alleninstitute.org/news/partnering-with-anthropic-to-supercharge-research-to-help-cure-disease",
        "role": "strategic_context",
        "title": "Allen Institute — Partnering with Anthropic to supercharge research and help cure disease"
      },
      {
        "url": "https://www.anthropic.com/news/anthropic-partners-with-allen-institute-and-howard-hughes-medical-institute",
        "role": "strategic_context",
        "title": "Anthropic — Partners with Allen Institute and HHMI to accelerate biological discovery"
      },
      {
        "url": "https://alleninstitute.org/person/andrew-hickl",
        "role": "persona_anchor",
        "title": "Allen Institute — Andrew Hickl person page"
      },
      {
        "url": "https://www.andyhickl.net/publications",
        "role": "research_background",
        "title": "Andy Hickl — publications index (NLP, QA, textual inference)"
      }
    ],
    "persona_id": "persona-andy-hickl",
    "authored_at": "2026-05-24T19:47:28.669024+00:00",
    "source_gap_ref": "knowledge_gap:ff660589-8438-4083-9973-a3c7aafc2956",
    "methods_summary": "Phase 1 — Extract trajectory artifact graphs for Kyle, Claire, Jerome, Kris from substrate using scidex.trajectory.list and scidex.links. Phase 2 — For each cross-persona link, verify: (a) provenance_chain non-empty, (b) reviewer handoff artifact exists and is complete, (c) refs_json contains ≥2 primary sources. Phase 3 — Score each trajectory: rigor_contract_rate = complete_reviews / total_artifacts; provenance_integrity = unbroken_chains / total_cross_links; citation_density = avg refs per artifact. Phase 4 — Compare scores across personas and against ARTIL design targets. Phase 5 — Record findings as benchmark artifact with baseline scores.",
    "expected_runtime": "2-3 agent cycles",
    "showcase_query_id": "ai-for-biology-throughput-rigor-loop"
  },
  "elo_score": "1000.0",
  "content_hash": "sha256:851934c6d383546d9db3806e5d3bfa6ed32b34023618a8e4f3a1f79879d118cb",
  "version_number": 1
}