Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Manage GitLab merge requests, issues, pipelines, and CI/CD from the terminal. Use when an agent needs to create MRs, manage issues, or monitor CI/CD pipelines on GitLab.
.claude/skills/composio-community-gitlab-cli-glab/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | -61% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -73% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 10% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -25% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | -26% | 0% |
MRs, issues, pipelines, and CI/CD management from the terminal.
bashbrew install glab glab auth login
bashglab mr list --output json glab mr create --title "Feature" --description "Details" glab issue list --output json glab issue create --title "Bug" --description "Details" glab ci list glab ci view <pipeline-id> glab repo clone owner/repo
--output json for structured outputGITLAB_TOKEN| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 5,271 | 2,657 | -50% | 1 | 1 | 0% | 741 | 575 | -22% | 0 | 0 | — |
case-02 | fail→pass | 23,860 | 4,789 | -80% | 1 | 1 | 0% | 2,210 | 870 | -61% | 0 | 0 | — |
case-03 | fail→pass | 16,697 | 4,071 | -76% | 1 | 1 | 0% | 2,220 | 592 | -73% | 0 | 0 | — |
case-04 | pass→pass | 5,882 | 3,783 | -36% | 1 | 1 | 0% | 948 | 705 | -26% | 0 | 0 | — |
case-05 | pass→pass | 7,563 | 2,475 | -67% | 1 | 1 | 0% | 1,119 | 530 | -53% | 0 | 0 | — |
case-06 | pass→pass | 7,897 | 1,971 | -75% | 1 | 1 | 0% | 1,136 | 448 | -61% | 0 | 0 | — |
case-07 | pass→pass | 12,024 | 3,861 | -68% | 1 | 1 | 0% | 1,422 | 532 | -63% | 0 | 0 | — |
case-13 | pass→pass | 14,378 | 2,285 | -84% | 1 | 1 | 0% | 1,128 | 469 | -58% | 0 | 0 | — |
case-08 | pass→pass | 7,843 | 4,037 | -49% | 1 | 1 | 0% | 1,383 | 651 | -53% | 0 | 0 | — |
case-09 | pass→pass | 3,376 | 1,611 | -52% | 1 | 1 | 0% | 334 | 378 | +13% | 0 | 0 | — |
case-10 | fail→pass | 4,090 | 2,586 | -37% | 1 | 1 | 0% | 501 | 550 | +10% | 0 | 0 | — |
case-11 | fail→pass | 9,952 | 7,384 | -26% | 1 | 1 | 0% | 1,494 | 1,116 | -25% | 0 | 0 | — |
case-12 | pass→pass | 10,963 | 3,981 | -64% | 1 | 1 | 0% | 1,100 | 534 | -51% | 0 | 0 | — |
case-14 | pass→pass | 3,386 | 2,405 | -29% | 1 | 1 | 0% | 513 | 411 | -20% | 0 | 0 | — |
case-15 | pass→pass | 7,058 | 2,243 | -68% | 1 | 1 | 0% | 486 | 562 | +16% | 0 | 0 | — |
case-16 | pass→pass | 5,636 | 17,030 | +202% | 1 | 1 | 0% | 857 | 610 | -29% | 0 | 0 | — |
case-17 | pass→pass | 7,833 | 3,027 | -61% | 1 | 1 | 0% | 1,321 | 556 | -58% | 0 | 0 | — |
case-18 | pass→pass | 12,483 | 2,877 | -77% | 1 | 1 | 0% | 1,286 | 517 | -60% | 0 | 0 | — |
case-19 | pass→pass | 9,924 | 2,496 | -75% | 1 | 1 | 0% | 1,423 | 388 | -73% | 0 | 0 | — |
case-20 | pass→pass | 9,015 | 7,696 | -15% | 1 | 1 | 0% | 1,623 | 1,511 | -7% | 0 | 0 | — |
case-21 | pass→pass | 6,271 | 2,612 | -58% | 1 | 1 | 0% | 909 | 581 | -36% | 0 | 0 | — |
case-22 | pass→pass | 6,525 | 3,510 | -46% | 1 | 1 | 0% | 1,123 | 752 | -33% | 0 | 0 | — |
case-23 | pass→pass | 12,232 | 12,621 | +3% | 1 | 1 | 0% | 2,101 | 2,055 | -2% | 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. The headline lift of +17 percentage points is the difference between those two pass rates over the 23 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.