A meaningful PR is not measured by lines changed. It is measured by clarity, testability, and respect for maintainer time — the same bar Btrust/Chaincode-style BOSS programs expect.
Rubric (score each 0–2)
| Row | 0 | 1 | 2 |
|---|---|---|---|
| Problem fit | Vague / wrong repo | Plausible issue | Confirmed with maintainers or clear GOOD FIRST ISSUE + skill match |
| Scope | Kitchen-sink | Slightly wide | Minimal diff that solves one stated problem |
| Correctness | Untested claim | Manual check only | Automated test, scripted repro, or documented verification steps |
| Communication | Drive-by dump | Basic description | Context, motivation, alternatives, and how to review |
| Citizenship | Ignores guide | Skims CONTRIBUTING | Follows CONTRIBUTING, style, CI, and review norms |
Ship bar: total ≥ 8/10 before you request maintainer review.
Evidence pack for Pull review
When you submit contribution progress on Pull, include:
- Issue / PR URLs
- Rubric self-score
- What you tested (commands, CI link, or Polar/regtest notes)
- Open questions for reviewers
Anti-patterns
- Drive-by refactors unrelated to the issue
- “LGTM” reviews with no reading
- Demanding merge timelines
- Mainnet experiments as proof