paper-sync

To check if your paper draft matches experiment outputs, reconcile manuscript claims against experiments/*.json and metrics outputs to verify consistency and flag unsupported statements.

Manuscript and experiment consistency verification is the process of cross-checking draft sections with README.md, CLAUDE.md, and docs/EXPERIMENTS.md to surface inconsistencies and ensure publication accuracy.

Yes, you need experiments/*.json files and metrics outputs to reconcile paper content, as these provide the underlying data required to verify alignment between reported results and manuscript claims.

Generate a revision-ready report for publication accuracy by applying an alignment workflow across manuscript artifacts and notebook outputs, which surfaces inconsistencies and produces structured revision suggestions.

The best way to align README.md and CLAUDE.md with experimental evidence is to cross-check these artifacts against experiments.json and notebook outputs to identify gaps and avoid stale or unsupported statements.

Run a paper and experiment consistency check before drafting, revising, or finishing the paper to ensure your manuscript does not contain stale claims or unsupported statements prior to publication.