Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Run AI-powered code review using the codex CLI. Use when the user asks to "codex review", "run codex review", or "review a commit with codex".
.claude/skills/tobihagemann-codex-review/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | -64% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -66% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -65% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -75% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -70% | 0% |
AI-powered code review via the codex CLI. Runs non-interactively.
bashcodex review --uncommitted
bashcodex review --base main codex review --base develop
bashcodex review --commit <sha> codex review --commit HEAD~1
Cannot be combined with --uncommitted, --base, or --commit.
bashcodex review "Focus on security issues and error handling"
--title to add context when reviewing feature branches or PRstimeout: 600000, do not set run_in_background)| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-04 | fail→pass | 9,274 | 3,841 | -59% | 1 | 1 | 0% | 1,341 | 483 | -64% | 0 | 0 | — |
case-01 | fail→fail | 5,887 | 4,138 | -30% | 1 | 1 | 0% | 901 | 331 | -63% | 0 | 0 | — |
case-02 | fail→fail | 6,966 | 4,793 | -31% | 1 | 1 | 0% | 1,154 | 380 | -67% | 0 | 0 | — |
case-03 | fail→fail | 9,668 | 7,739 | -20% | 1 | 1 | 0% | 662 | 368 | -44% | 0 | 0 | — |
case-05 | fail→pass | 7,314 | 1,843 | -75% | 1 | 1 | 0% | 1,242 | 421 | -66% | 0 | 0 | — |
case-06 | fail→pass | 8,378 | 2,043 | -76% | 1 | 1 | 0% | 1,383 | 483 | -65% | 0 | 0 | — |
case-07 | fail→fail | 4,865 | 4,577 | -6% | 1 | 1 | 0% | 718 | 347 | -52% | 0 | 0 | — |
case-08 | fail→fail | 3,999 | 5,064 | +27% | 1 | 1 | 0% | 626 | 365 | -42% | 0 | 0 | — |
case-09 | fail→fail | 12,183 | 10,277 | -16% | 1 | 1 | 0% | 2,133 | 1,040 | -51% | 0 | 0 | — |
case-10 | fail→pass | 9,632 | 1,627 | -83% | 1 | 1 | 0% | 1,638 | 405 | -75% | 0 | 0 | — |
case-11 | fail→pass | 10,308 | 2,112 | -80% | 1 | 1 | 0% | 1,550 | 469 | -70% | 0 | 0 | — |
case-12 | pass→pass | 11,364 | 2,040 | -82% | 1 | 1 | 0% | 1,772 | 482 | -73% | 0 | 0 | — |
case-13 | fail→pass | 10,104 | 2,076 | -79% | 1 | 1 | 0% | 1,786 | 477 | -73% | 0 | 0 | — |
case-14 | fail→fail | 6,274 | 5,071 | -19% | 1 | 1 | 0% | 1,003 | 365 | -64% | 0 | 0 | — |
case-15 | fail→pass | 9,507 | 2,061 | -78% | 1 | 1 | 0% | 1,578 | 475 | -70% | 0 | 0 | — |
case-16 | fail→pass | 9,026 | 1,709 | -81% | 1 | 1 | 0% | 1,475 | 378 | -74% | 0 | 0 | — |
case-17 | fail→fail | 3,632 | 4,574 | +26% | 1 | 1 | 0% | 520 | 401 | -23% | 0 | 0 | — |
case-18 | fail→pass | 11,538 | 2,221 | -81% | 1 | 1 | 0% | 1,945 | 452 | -77% | 0 | 0 | — |
case-19 | fail→fail | 18,977 | 7,246 | -62% | 1 | 1 | 0% | 1,424 | 336 | -76% | 0 | 0 | — |
case-20 | pass→pass | 3,053 | 2,368 | -22% | 1 | 1 | 0% | 495 | 577 | +17% | 0 | 0 | — |
case-21 | pass→pass | 6,289 | 4,264 | -32% | 1 | 1 | 0% | 1,075 | 894 | -17% | 0 | 0 | — |
case-22 | pass→pass | 2,532 | 1,890 | -25% | 1 | 1 | 0% | 370 | 417 | +13% | 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 14 counted toward the lift figure. The other 8 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 14 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.