Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Gate a Sentry feature behind a FlagPole feature flag. Use when adding a feature flag, registering a flag in temporary.py, checking a flag from Python or the frontend, enabling a flag in tests, or asking where FlagPole rollout config lives. Trigger on "add a feature flag", "gate this behind a flag", "register a flag", "features.has", "api_expose", "OrganizationFeature", "ProjectFeature", "FlagPole".
.claude/skills/getsentry-feature-flags/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -37% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -40% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -56% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -69% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -60% | 0% |
New features should be gated behind a feature flag.
src/sentry/features/temporary.py:python manager.add("organizations:my-feature", OrganizationFeature, FeatureHandlerStrategy.FLAGPOLE, api_expose=True)
Use api_expose=True if the frontend needs to check the flag. Use ProjectFeature and a projects: prefix for project-scoped flags.
python if features.has("organizations:my-feature", organization, actor=user):
api_expose=True):typescript organization.features.includes('my-feature');
python with self.feature("organizations:my-feature"): ...
sentry-options-automator repo, not here.remove-option-or-flag skill.See https://develop.sentry.dev/feature-flags/ for full docs.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 20,861 | 13,018 | -38% | 1 | 1 | 0% | 3,238 | 2,026 | -37% | 0 | 0 | — |
case-02 | fail→pass | 17,235 | 10,266 | -40% | 1 | 1 | 0% | 2,223 | 1,331 | -40% | 0 | 0 | — |
case-03 | fail→pass | 22,733 | 9,325 | -59% | 1 | 1 | 0% | 2,836 | 1,243 | -56% | 0 | 0 | — |
case-04 | pass→pass | 13,431 | 10,139 | -25% | 1 | 1 | 0% | 1,625 | 1,161 | -29% | 0 | 0 | — |
case-05 | pass→pass | 12,505 | 7,665 | -39% | 1 | 1 | 0% | 1,322 | 678 | -49% | 0 | 0 | — |
case-06 | fail→pass | 18,258 | 7,760 | -57% | 1 | 1 | 0% | 2,583 | 796 | -69% | 0 | 0 | — |
case-07 | pass→pass | 12,666 | 6,835 | -46% | 1 | 1 | 0% | 1,342 | 632 | -53% | 0 | 0 | — |
case-08 | pass→pass | 12,375 | 7,003 | -43% | 1 | 1 | 0% | 1,478 | 668 | -55% | 0 | 0 | — |
case-09 | fail→pass | 13,676 | 6,882 | -50% | 1 | 1 | 0% | 1,508 | 609 | -60% | 0 | 0 | — |
case-10 | pass→pass | 10,637 | 7,521 | -29% | 1 | 1 | 0% | 1,046 | 750 | -28% | 0 | 0 | — |
case-11 | pass→pass | 11,768 | 6,445 | -45% | 1 | 1 | 0% | 1,198 | 500 | -58% | 0 | 0 | — |
case-12 | pass→pass | 10,574 | 7,009 | -34% | 1 | 1 | 0% | 953 | 623 | -35% | 0 | 0 | — |
case-13 | pass→pass | 10,019 | 1,849 | -82% | 1 | 1 | 0% | 951 | 644 | -32% | 0 | 0 | — |
case-14 | pass→pass | 9,047 | 6,927 | -23% | 1 | 1 | 0% | 741 | 597 | -19% | 0 | 0 | — |
case-15 | fail→pass | 11,279 | 6,550 | -42% | 1 | 1 | 0% | 926 | 512 | -45% | 0 | 0 | — |
case-16 | pass→pass | 6,917 | 3,261 | -53% | 1 | 1 | 0% | 1,334 | 873 | -35% | 0 | 0 | — |
case-17 | fail→pass | 14,645 | 11,986 | -18% | 1 | 1 | 0% | 2,396 | 1,576 | -34% | 0 | 0 | — |
case-18 | pass→pass | 16,833 | 7,947 | -53% | 1 | 1 | 0% | 2,056 | 819 | -60% | 0 | 0 | — |
case-19 | fail→pass | 18,613 | 2,002 | -89% | 1 | 1 | 0% | 2,730 | 679 | -75% | 0 | 0 | — |
case-20 | pass→pass | 22,563 | 8,998 | -60% | 1 | 1 | 0% | 2,725 | 1,745 | -36% | 0 | 0 | — |
case-21 | fail→pass | 19,923 | 8,148 | -59% | 1 | 1 | 0% | 2,736 | 1,797 | -34% | 0 | 0 | — |
case-22 | fail→fail | 18,702 | 9,096 | -51% | 1 | 1 | 0% | 2,459 | 2,180 | -11% | 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 +41 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.
| Model | Method | Date | Lift |
|---|---|---|---|
| gemini-3.6-flash | verified | 8/28/2026 | +41% |
Other measured skills in the registry, with their headline benchmark lift.