Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Write a CONTRIBUTING guide that helps people contribute to an open-source project without friction. Use when asked to write a CONTRIBUTING.md, set up contribution guidelines, or make a repo welcoming to contributors. Produces a clear guide: how to set up, the contribution workflow, standards, PR expectations, and how to get help — lowering the barrier to a first PR.
.claude/skills/mohitagw15856-contributor-guide/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 11% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 24% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 52% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 53% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | 31% | 0% |
Most would-be contributors give up at setup friction or unclear expectations. A good CONTRIBUTING.md removes the guesswork: how to get the project running, how to propose a change, what a mergeable PR looks like, and where to ask. This skill writes that guide — welcoming, specific, and aimed at getting someone to a successful first PR.
Ask for these only if they aren't already provided:
A CONTRIBUTING.md:
A warm one-liner: contributions are welcome, here's how to make it smooth.
Ways to contribute — issues, docs, code, triage — not everyone writes code.
Development setup
# clone, install, run, test — the exact commands…so a contributor can get the project running and tests passing locally.
Finding something to work on — point to good first issue / help wanted; ask people to comment before starting larger work.
Making a change (the workflow)
Standards — formatting/linting, test expectations, commit/PR conventions, the Code of Conduct link.
What happens next — who reviews, rough turnaround, how feedback works.
Getting help — where to ask (Discussions, chat, issue) — make it explicitly OK to ask.
Open-source contribution best practices (clear setup, defined workflow, good-first-issues, welcoming tone, CoC).
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 12,714 | 10,145 | -20% | 1 | 1 | 0% | 2,242 | 2,493 | +11% | 0 | 0 | — |
case-02 | fail→pass | 13,181 | 22,266 | +69% | 1 | 1 | 0% | 2,512 | 3,112 | +24% | 0 | 0 | — |
case-03 | pass→pass | 78,301 | 21,580 | -72% | 1 | 1 | 0% | 3,476 | 4,561 | +31% | 0 | 0 | — |
case-04 | pass→pass | 12,485 | 11,642 | -7% | 1 | 1 | 0% | 2,299 | 2,454 | +7% | 0 | 0 | — |
case-05 | pass→pass | 8,324 | 9,927 | +19% | 1 | 1 | 0% | 1,314 | 2,244 | +71% | 0 | 0 | — |
case-06 | fail→pass | 9,804 | 11,188 | +14% | 1 | 1 | 0% | 1,636 | 2,493 | +52% | 0 | 0 | — |
case-07 | fail→fail | 9,614 | 15,805 | +64% | 1 | 1 | 0% | 1,689 | 3,127 | +85% | 0 | 0 | — |
case-08 | fail→fail | 10,518 | 9,976 | -5% | 1 | 1 | 0% | 1,517 | 2,472 | +63% | 0 | 0 | — |
case-09 | fail→fail | 11,532 | 12,886 | +12% | 1 | 1 | 0% | 2,068 | 3,051 | +48% | 0 | 0 | — |
case-10 | fail→fail | 21,620 | 14,093 | -35% | 1 | 1 | 0% | 2,038 | 2,938 | +44% | 0 | 0 | — |
case-11 | fail→pass | 9,829 | 10,949 | +11% | 1 | 1 | 0% | 1,641 | 2,507 | +53% | 0 | 0 | — |
case-12 | fail→fail | 9,783 | 12,744 | +30% | 1 | 1 | 0% | 1,787 | 2,746 | +54% | 0 | 0 | — |
case-13 | fail→fail | 4,475 | 7,112 | +59% | 1 | 1 | 0% | 757 | 1,999 | +164% | 0 | 0 | — |
case-14 | fail→fail | 8,699 | 20,688 | +138% | 1 | 1 | 0% | 1,416 | 2,442 | +72% | 0 | 0 | — |
case-15 | pass→pass | 14,061 | 14,958 | +6% | 1 | 1 | 0% | 2,397 | 3,218 | +34% | 0 | 0 | — |
case-16 | pass→pass | 11,140 | 31,137 | +180% | 1 | 1 | 0% | 1,793 | 2,892 | +61% | 0 | 0 | — |
case-17 | fail→fail | 7,187 | 10,479 | +46% | 1 | 1 | 0% | 1,194 | 2,685 | +125% | 0 | 0 | — |
case-18 | fail→fail | 9,062 | 10,550 | +16% | 1 | 1 | 0% | 1,709 | 2,399 | +40% | 0 | 0 | — |
case-19 | fail→fail | 14,065 | 10,964 | -22% | 1 | 1 | 0% | 2,554 | 2,538 | -1% | 0 | 0 | — |
case-20 | fail→fail | 7,061 | 9,052 | +28% | 1 | 1 | 0% | 1,202 | 2,354 | +96% | 0 | 0 | — |
case-21 | fail→fail | 6,230 | 9,166 | +47% | 1 | 1 | 0% | 1,185 | 2,451 | +107% | 0 | 0 | — |
case-22 | pass→pass | 13,767 | 12,365 | -10% | 1 | 1 | 0% | 2,151 | 2,966 | +38% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +18 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.