| name | robustness-checker |
| description | Design and run risk-targeted robustness, sensitivity, error, and baseline checks for an approved mathematical model, emitting compact machine evidence in lean mode and a final report in submission mode. |
Purpose
Test the claims most likely to fail. Choose checks from the model's assumptions and decision risks rather than filling a generic checklist.
Preconditions
- Approved main and usable baseline executed.
- Run summary, method card, probe summary, and relevant outputs exist.
- Claim or decision to be tested is known.
Workflow
- Identify load-bearing assumptions and claims.
- Select applicable checks:
- parameter or weight perturbation;
- alternate split or resampling;
- seed stability;
- outlier/missing-data treatment;
- constraint/capacity perturbation;
- baseline comparison;
- output concentration/rank stability;
- error and uncertainty analysis.
- State perturbation ranges and why they are meaningful before interpreting results.
- Run checks with fixed seeds where stochastic.
- Save compact metrics to:
robustness/Qx/qx_robustness_summary.json
- In
submission, also save:
robustness/Qx/qx_robustness_report.md
- If the stability verdict affects method continuation or claim scope, invoke one choice card and log the human answer in
qx_decisions.jsonl.
Summary Contract
Record:
- tested claim/assumption;
- input and result source paths;
- perturbation;
- metric and threshold if predeclared;
- observed value;
- status
PASS, CONDITIONAL, or FAIL;
- limitation;
- fallback-trigger relevance.
Rules
- Do not run irrelevant checks merely to reach a count.
- Do not invent a threshold after seeing the result without labeling it exploratory.
- Do not convert stability metrics into the human confidence verdict.
- Do not create
robustness-checker_modeler_decision.md.
- A failed robustness check is evidence for adjust/fallback/claim downgrade, not permission for AI to decide.
Verification
- Every major final claim has a supporting check or explicit limitation.
- Perturbations are justified and reproducible.
- Baseline and main comparisons remain metric-compatible.
- Concentration/degeneracy risks are revisited when relevant.
- Submission report sources its numbers from the summary and experiment artifacts.