{
  "schema_version": "1.0",
  "version": "0.2",
  "created_at": "2026-10-23",
  "forecast_id": "A-001",
  "assessment_type": "internal_editorial_pre_review_NOT_independent",
  "external_reviewers_appointed": 0,
  "external_signoffs": 0,
  "threshold_for_independent_approval": 2,
  "result": "HOLD",
  "official_probability": null,
  "previous_working_model_display_percent": 85,
  "corrected_working_model_display_percent": 86,
  "correction_explanation": "0.95 x 0.90 = 0.855 = 85.5%; standard whole-percent rounding -> 86%",
  "sensitivity_percent": [
    64,
    96
  ],
  "sensitivity_is_statistical_confidence_interval": false,
  "independent_calibration_validated": false,
  "review_verdict": "The METR task-horizon methodology is real, but the two 95% and 90% conditional scenario inputs are unsourced editorial judgments. An 80%-task-success horizon is not 80% probability of future attainment.",
  "verified_source_scope": [
    "METR defines its time horizon as a human-expert task duration at 50% or 80% modeled success; it is not autonomous operating time.",
    "METR public time-horizon dashboard says it is no longer actively updated and its task distribution mainly covers software/ML/cybersecurity."
  ],
  "blocking_findings": [
    {
      "code": "AGENT-EVAL",
      "severity": "critical",
      "task": "Select and preregister an independent, stable 80%-reliability benchmark or validated METR successor, including task mix, sample size and rules for post-2026 discontinuation."
    },
    {
      "code": "AGENT-REFCLASS",
      "severity": "critical",
      "task": "Reconstruct comparable historical 80% task-horizon observations with time stamps and uncertainty; audit the trend using rolling-origin forecasting and holdouts."
    },
    {
      "code": "AGENT-PUB",
      "severity": "high",
      "task": "Fix agent public-availability date, replication conditions, missing-data treatment and 'no evaluator by deadline' unresolved rule."
    },
    {
      "code": "AGENT-SIGN",
      "severity": "critical",
      "task": "Obtain two named external conflict-screened independent signoffs and a third-party timestamp before any official forecast is published."
    }
  ],
  "external_roles_needed": [
    "Independent technical AI agent evaluator familiar with METR task-horizon suite",
    "Forecasting statistician with out-of-sample backtesting expertise"
  ],
  "primary_sources": [
    "https://metr.org/time-horizons/",
    "https://metr.org/blog/2026-05-19-frontier-risk-report/",
    "https://www.annualreviews.org/content/journals/10.1146/annurev-statistics-042424-050626"
  ],
  "lock_status": "NOT_READY_TO_LOCK",
  "scoring_eligible": false,
  "public_reviewer_contact": "https://cortist.com/contact/",
  "no_retroactive_lock": true,
  "versioned_prior_workpaper": "/data/forecasts/calibration/a-001-working-20261023.json"
}