Approximate counts are welcome; say how you counted them in your notes. For lines of code, exclude blank lines, comments, generated and vendored code. Human-reviewed requests have a review from someone other than the author; exclude bots.

Choose a few that tell the story. Test-to-source ratio is a file ratio, not test coverage.

Describe what exists, such as architecture notes, review discussions, runbooks, schemas or issue history. Don’t paste the actual documents or personal data.

Ran the repo provenance check? Paste all of goodwattle-report.json. We keep its counts, licence families and hash, not file paths. A link from the codebase value estimator is welcome too. Both are recorded as your own results, not verified evidence.