Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Goal-driven development loop — define objective, write rules with key-results, verify visually, sync to issue tracker.
.claude/skills/x-cmd-devloop/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -2% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -41% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -38% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -55% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -58% | 0% |
x rule init :ror # or project-specific ruleset
x env use agent-browser # for visual verificationgoal → rule.yml → code → verify → issueBefore coding, create a rule.yml for the task:
yamlgoal: "Fix English version showing Chinese tags in Features section" keyresults: - kr-1: "English mode Features tags are all English" - kr-2: "Chinese mode Features tags remain Chinese" - kr-3: "Build passes with no regressions" rules: - id: kr-1-verify name: english-tags-no-chinese apply: "src/components/Features.tsx" level: error desc: - When language=en, all details[] items must be English - No Chinese characters in any tag badge in English mode
Rule fields: id, name, apply, level, desc, tldr. See x rule -h.
MUST include goal and keyresults — the two mandatory fields for every dev task.
agent-browser open <url> --session <proj> --headed
agent-browser --session <proj> screenshot /tmp/before.png
# ... make changes ...
agent-browser --session <proj> screenshot /tmp/after.pngMUST screenshot before and after. Compare against intent:
x rule scan <files> # fast (~1 min), for inner loop
x rule check <files> # thorough (~10 min)
x rule audit <files> # full report (~30 min)See issue/SKILL.md. Minimum: post goal, key-results, rule.yml, before/after screenshots to the issue.
x rule scan — Quick check during iterationx rule check — Full compliance before commitagent-browser screenshot — Visual proof before/aftergh issue create/comment — Sync goal + rule.yml to tracker| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 14,024 | 8,323 | -41% | 1 | 1 | 0% | 2,307 | 2,258 | -2% | 0 | 0 | — |
case-02 | fail→fail | 8,700 | 6,465 | -26% | 1 | 1 | 0% | 327 | 976 | +198% | 0 | 0 | — |
case-03 | fail→fail | 10,894 | 8,757 | -20% | 1 | 1 | 0% | 174 | 1,016 | +484% | 0 | 0 | — |
case-04 | pass→pass | 11,898 | 4,529 | -62% | 1 | 1 | 0% | 1,896 | 1,358 | -28% | 0 | 0 | — |
case-05 | fail→pass | 11,420 | 2,711 | -76% | 1 | 1 | 0% | 1,772 | 1,037 | -41% | 0 | 0 | — |
case-06 | fail→pass | 14,608 | 2,759 | -81% | 1 | 1 | 0% | 1,669 | 1,039 | -38% | 0 | 0 | — |
case-07 | fail→pass | 18,195 | 3,468 | -81% | 1 | 1 | 0% | 2,667 | 1,202 | -55% | 0 | 0 | — |
case-08 | fail→pass | 31,923 | 2,269 | -93% | 1 | 1 | 0% | 2,280 | 955 | -58% | 0 | 0 | — |
case-09 | fail→pass | 12,216 | 2,202 | -82% | 1 | 1 | 0% | 1,984 | 925 | -53% | 0 | 0 | — |
case-10 | fail→pass | 7,715 | 3,114 | -60% | 1 | 1 | 0% | 1,293 | 1,080 | -16% | 0 | 0 | — |
case-11 | pass→pass | 8,851 | 2,970 | -66% | 1 | 1 | 0% | 1,203 | 1,116 | -7% | 0 | 0 | — |
case-12 | fail→pass | 11,596 | 3,233 | -72% | 1 | 1 | 0% | 1,713 | 1,084 | -37% | 0 | 0 | — |
case-13 | fail→pass | 9,939 | 5,702 | -43% | 1 | 1 | 0% | 1,100 | 1,638 | +49% | 0 | 0 | — |
case-14 | fail→pass | 11,589 | 3,571 | -69% | 1 | 1 | 0% | 1,713 | 1,181 | -31% | 0 | 0 | — |
case-15 | fail→pass | 10,777 | 4,218 | -61% | 1 | 1 | 0% | 1,723 | 1,394 | -19% | 0 | 0 | — |
case-16 | fail→pass | 11,954 | 2,571 | -78% | 1 | 1 | 0% | 1,628 | 1,036 | -36% | 0 | 0 | — |
case-17 | pass→pass | 5,578 | 2,627 | -53% | 1 | 1 | 0% | 792 | 1,068 | +35% | 0 | 0 | — |
case-18 | pass→pass | 37,279 | 9,061 | -76% | 1 | 1 | 0% | 6,021 | 1,121 | -81% | 0 | 0 | — |
case-19 | fail→pass | 7,803 | 2,479 | -68% | 1 | 1 | 0% | 1,131 | 956 | -15% | 0 | 0 | — |
case-20 | pass→pass | 7,324 | 46,094 | +529% | 1 | 1 | 0% | 928 | 1,176 | +27% | 0 | 0 | — |
case-21 | pass→pass | 15,634 | 6,718 | -57% | 1 | 1 | 0% | 1,684 | 1,979 | +18% | 0 | 0 | — |
case-22 | pass→fail | 10,336 | 8,795 | -15% | 1 | 1 | 0% | 1,818 | 1,804 | -1% | 0 | 0 | — |
case-23 | pass→pass | 11,251 | 10,874 | -3% | 1 | 1 | 0% | 1,992 | 2,381 | +20% | 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. 23 cases were attempted, and 20 counted toward the lift figure. The other 3 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 +52 percentage points is the difference between those two pass rates over the 20 comparable cases. 1 case got worse with the skill loaded, and it is 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.