hypothesis-simulation-review

hypothesis-simulation-review

Goal:

• Simulate the hypothesis mechanism and identify failure scenarios.

Inputs:

• research_plan/RESEARCH_PLAN.json
• hypotheses/<id>/HYPOTHESIS.json
• optional prior review artifacts for calibration

Outputs:

• hypotheses/<id>/REVIEW/SIMULATION_REVIEW.json

Context Loading:

• Read research_plan/RESEARCH_PLAN.json.
• Read hypotheses/<id>/HYPOTHESIS.json.
• Use prior reviews only as hints for likely weak spots; the simulation should still be an independent walk-through of the mechanism and validation path.

Execution Prompt Contract:

• System Intent:
  • You are mentally simulating the hypothesis mechanism and proposed validation path step by step.
• Required Reasoning Focus:
  • Break the mechanism into an ordered process.
  • Identify where the process could fail and why.
  • Distinguish inherent failures from contingent implementation issues where possible.
  • Prioritize likely and fundamental failure scenarios over edge cases.
• Do Not Do:
  • Do not generate abstract critique without a concrete simulated step.
  • Do not flood the artifact with low-value edge cases.
  • Do not replace the structured review with narrative prose.
• Review Quality Floor:
  • A status = completed simulation review must include concrete mechanism or validation steps and concrete failure scenarios.
  • Steps must name hypothesis-specific materials, mechanisms, measurements, conditions, controls, or decision points.
  • Do not use placeholder simulation steps such as Synthesize evolved catalyst, Characterize structure, Test activity, or Benchmark against parent as substantive review content.
• Output Shape:
  • Produce the exact SimulationReviewContract from packages/agent_contracts/review.py with:
    • steps
    • failure_scenarios
  • Keep each step and failure scenario concise and concrete.
  • Each item in steps must be a plain string, not an object with nested keys such as step, description, or expected_outcome.
  • Each item in failure_scenarios must be a plain string, not an object with nested keys such as scenario, likelihood, or mitigation.

Execution Steps:

1. Open skills/shared-references/schema-index.md, then read packages/agent_contracts/review.py and confirm the exact SimulationReviewContract shape before writing SIMULATION_REVIEW.json.
2. Read the research plan and hypothesis.
3. Simulate the key mechanism or validation path step by step.
4. Record the major steps.
5. Identify likely and fundamental failure scenarios.
6. Write hypotheses/<id>/REVIEW/SIMULATION_REVIEW.json.
7. Validate before declaring completion.

Artifact Rules:

• SIMULATION_REVIEW.json should focus on the highest-signal failure modes.
• The artifact should remain short enough for downstream summarization but concrete enough to guide revision.

Completion Rule:

• This skill is complete only when SIMULATION_REVIEW.json exists and is valid for downstream synthesis.