Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Auto-fix lint and format issues across all affected services
.claude/skills/nudgebee-fix-lint-format-issues/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-09 | ✗→✓ | ▲ Improved | -6% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -27% | 0% |
| case-12 | ✗→✓ | ▲ Improved | -39% | 0% |
| case-13 | ✗→✓ | ▲ Improved | -44% | 0% |
| case-14 | ✗→✓ | ▲ Improved | -28% | 0% |
Auto-fix linting and formatting issues for affected services. Optional argument: $ARGUMENTS (specific service name, or --all to fix everything).
If $ARGUMENTS is a service name: fix only that service. If $ARGUMENTS is --all: fix all services in the repo. If $ARGUMENTS is empty: detect affected services from changed files:
bash# Changes vs main (committed + uncommitted) git diff --name-only main...HEAD git diff --name-only git diff --name-only --cached
Combine and deduplicate, then map to services.
Run the appropriate fix command for each affected service. Execute independent services in parallel where possible.
| Service | Fix command | Working directory | |---|---|---| | api-server | make fmt | api-server/services/ | | ticket-server | make fmt | ticket-server/ | | cloud-collector | make fmt | collector-server/cloud-collector/ | | relay-server | make fmt | collector-server/k8s-collector/relay-server/ | | k8s-collector-app | poetry run black . | collector-server/k8s-collector/app/ | | code-analysis | make fmt | llm/code-analysis/ | | llm-server | make fmt | llm/llm-server/ | | rag-server | poetry run black . | llm/rag-server/ | | benchmark | poetry run black . | llm/benchmark/ | | ml-k8s-server | make fmt | ml-k8s-server/ | | auto-pilot | poetry run black . | auto-pilot/ | | auto-pilot-sidecar | poetry run black . | auto-pilot/sidecar/ | | notifications-server | poetry run black . | notifications-server/ | | frontend | npm run lint2:fix | app/ |
After fixing, re-run the check (not fix) command for each service to confirm everything passes:
| Service type | Verify command | |---|---| | Go (with Makefile) | make lint | | Python (with Makefile) | make lint | | Python (no Makefile) | poetry run black --check . && poetry run flake8 . | | TypeScript | npm run lint2 |
## Fix Results
| Service | Fix | Verify | Status |
|---------|-----|--------|--------|
| api-server | make fmt | make lint | PASS/FAIL |
| app | npm run lint2:fix | npm run lint2 | PASS/FAIL |
{N} services fixed, {P} passing, {F} still failingIf any service still fails after auto-fix, show the remaining errors. These likely need manual intervention (e.g., flake8 issues that black can't fix, type errors from mypy).
bashgit diff --name-only
Tell the user what files were modified by the auto-fix so they can review before staging.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-09 | fail→pass | 7,771 | 2,545 | -67% | 1 | 1 | 0% | 1,245 | 1,169 | -6% | 0 | 0 | — |
case-06 | pass→pass | 9,771 | 9,327 | -5% | 1 | 1 | 0% | 1,643 | 2,362 | +44% | 0 | 0 | — |
case-08 | fail→fail | 9,083 | 8,414 | -7% | 1 | 1 | 0% | 1,351 | 1,180 | -13% | 0 | 0 | — |
case-07 | fail→pass | 9,355 | 10,185 | +9% | 1 | 1 | 0% | 1,601 | 1,171 | -27% | 0 | 0 | — |
case-01 | fail→fail | 3,186 | 4,851 | +52% | 1 | 1 | 0% | 343 | 1,032 | +201% | 0 | 0 | — |
case-02 | fail→fail | 3,416 | 14,296 | +319% | 1 | 1 | 0% | 164 | 1,000 | +510% | 0 | 0 | — |
case-03 | fail→fail | 3,820 | 11,244 | +194% | 1 | 1 | 0% | 258 | 1,482 | +474% | 0 | 0 | — |
case-04 | pass→fail | 19,533 | 5,135 | -74% | 1 | 1 | 0% | 1,911 | 1,026 | -46% | 0 | 0 | — |
case-05 | pass→fail | 10,058 | 8,862 | -12% | 1 | 1 | 0% | 1,788 | 2,197 | +23% | 0 | 0 | — |
case-10 | fail→fail | 21,084 | 1,618 | -92% | 1 | 1 | 0% | 3,435 | 1,053 | -69% | 0 | 0 | — |
case-11 | fail→fail | 7,574 | 2,832 | -63% | 1 | 1 | 0% | 1,126 | 1,251 | +11% | 0 | 0 | — |
case-12 | fail→pass | 10,945 | 2,132 | -81% | 1 | 1 | 0% | 1,845 | 1,122 | -39% | 0 | 0 | — |
case-13 | fail→pass | 11,078 | 1,619 | -85% | 1 | 1 | 0% | 1,796 | 1,009 | -44% | 0 | 0 | — |
case-14 | fail→pass | 10,082 | 2,581 | -74% | 1 | 1 | 0% | 1,633 | 1,180 | -28% | 0 | 0 | — |
case-15 | fail→fail | 6,276 | 2,494 | -60% | 1 | 1 | 0% | 942 | 1,130 | +20% | 0 | 0 | — |
case-16 | fail→pass | 9,241 | 2,307 | -75% | 1 | 1 | 0% | 1,418 | 1,133 | -20% | 0 | 0 | — |
case-17 | fail→pass | 6,465 | 2,060 | -68% | 1 | 1 | 0% | 982 | 1,070 | +9% | 0 | 0 | — |
case-18 | fail→pass | 7,885 | 2,748 | -65% | 1 | 1 | 0% | 1,371 | 1,186 | -13% | 0 | 0 | — |
case-19 | fail→pass | 8,896 | 2,053 | -77% | 1 | 1 | 0% | 1,469 | 1,082 | -26% | 0 | 0 | — |
case-20 | fail→pass | 10,575 | 2,738 | -74% | 1 | 1 | 0% | 1,855 | 1,160 | -37% | 0 | 0 | — |
case-21 | fail→fail | 12,089 | 1,617 | -87% | 1 | 1 | 0% | 1,864 | 1,009 | -46% | 0 | 0 | — |
case-22 | fail→pass | 5,091 | 1,440 | -72% | 1 | 1 | 0% | 865 | 989 | +14% | 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, and 18 counted toward the lift figure. The other 4 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +41 percentage points is the difference between those two pass rates over the 18 comparable cases. 2 cases got worse with the skill loaded, and they are included in that figure.
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.