Skip to main content

cumcm-independent-review

Independently review a CUMCM computation package before its claims enter validation or paper writing. Use only inside a generated independent-review package.

Zur Installation springen

Quellinformationen

Repository
Lucasuiii/cumcm-codex-workflow
Letzte Quellaktivität
31. August 2026 um 13:24
Erkannte Sprache von SKILL.md
Englisch
Sterne
0
Forks
0

Installationsoptionen

Standardmäßig ist der Prompt ausgewählt, der zuerst die Quelle prüft. Sie können zu einem direkten Befehl wechseln oder eine lokale Kopie herunterladen.

Quelldateien prüfen

Lesen Sie SKILL.md und alle von SkillsMP angezeigten Begleitdateien, bevor Sie sich für eine Installation entscheiden.

Datei-Explorer
2 Dateien

SKILL.md wird angezeigt

SKILL.md
Quellanweisungen · Schreibgeschützte Vorschau
name
cumcm-independent-review
description
Independently review a CUMCM computation package before its claims enter validation or paper writing. Use only inside a generated independent-review package.
# CUMCM Independent Review Review this package without consulting the originating conversation. Treat every supplied conclusion as untrusted and reconstruct the task from the official inputs, contracts, code, run records, and executed outputs. ## Boundaries - Work read-only inside the package. - Do not search for missing official materials. Report them as missing. - Do not edit, rerun, or replace the preserved execution unless the user separately authorizes a reproduction run. - File existence and successful execution do not prove that the model answers the official question. - Give exact file, formula, code, or numerical locations for every finding. - Preserve negative and inconclusive findings verbatim. ## Review order 1. Read `REVIEW_REQUEST.md` and `materials/problem/SOURCE_MANIFEST.json`. 2. Reconstruct each subproblem from the supplied official files and `PROBLEM_FACTS.json`. 3. Compare the official request with the model objective, variables, constraints, assumptions, and cross-question dependencies. 4. Inspect computation entry points, run manifests, executed outputs, and result locators. 5. Challenge at least these failure classes: - task or target misunderstood; - upper/lower bound or optimization direction reversed; - a quantity counted twice; - unsupported extrapolation; - an observed variable omitted without justification; - cross-question contradiction; - code and mathematical formulation disagree; - numerical output violates units, bounds, conservation, or official constraints. 6. Write the raw review to `INDEPENDENT_REVIEW_RAW.md` and the structured result to `INDEPENDENT_REVIEW_RESULT.json` using the supplied schema example. ## Verdict - `accepted`: no fatal or unresolved major defect was found in the reviewed scope. - `revision_requested`: at least one fatal or unresolved major defect requires returning to model design or computation. - `inconclusive`: required material is missing or the available evidence cannot support a decision. State the reviewer, model if applicable, task reference, whether this is a different conversation, and the appropriate independence grade. Never describe a same-context review as independent.
Auf GitHub ansehen