Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Feature flag configuration and rollout planning for controlled releases
.claude/skills/a5c-ai-feature-flagging/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | -12% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -18% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 27% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 43% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 100% | 0% |
Specialized skill for feature flag configuration and rollout planning. Enables product teams to plan and manage controlled feature releases with proper targeting, monitoring, and rollback strategies.
This skill integrates with the following processes:
product-launch-gtm.js - Feature flag rollout for launchesbeta-program.js - Beta feature flaggingconversion-funnel-analysis.js - A/B test flag managementjson{ "type": "object", "properties": { "mode": { "type": "string", "enum": ["design", "rollout", "audit", "deprecate"], "description": "Operation mode" }, "feature": { "type": "object", "properties": { "name": { "type": "string" }, "description": { "type": "string" }, "type": { "type": "string", "enum": ["release", "experiment", "ops", "permission"] }, "owner": { "type": "string" }, "impactLevel": { "type": "string", "enum": ["low", "medium", "high", "critical"] } } }, "rolloutStrategy": { "type": "object", "properties": { "type": { "type": "string", "enum": ["percentage", "segment", "geographic", "gradual"] }, "stages": { "type": "array", "items": { "type": "object" } }, "criteria": { "type": "object" } } }, "existingFlags": { "type": "array", "items": { "type": "object" }, "description": "Existing flags for audit or cleanup" } }, "required": ["mode"] }
json{ "type": "object", "properties": { "flagSpec": { "type": "object", "properties": { "key": { "type": "string" }, "name": { "type": "string" }, "description": { "type": "string" }, "type": { "type": "string" }, "defaultValue": { "type": "boolean" }, "variations": { "type": "array", "items": { "type": "object" } }, "targetingRules": { "type": "array", "items": { "type": "object" } }, "prerequisites": { "type": "array", "items": { "type": "string" } } } }, "rolloutPlan": { "type": "object", "properties": { "stages": { "type": "array", "items": { "type": "object", "properties": { "stage": { "type": "string" }, "percentage": { "type": "number" }, "targeting": { "type": "object" }, "duration": { "type": "string" }, "successCriteria": { "type": "array", "items": { "type": "string" } }, "rollbackTriggers": { "type": "array", "items": { "type": "string" } } } } }, "metrics": { "type": "array", "items": { "type": "string" } }, "alerts": { "type": "array", "items": { "type": "object" } } } }, "killSwitch": { "type": "object", "properties": { "procedure": { "type": "array", "items": { "type": "string" } }, "owner": { "type": "string" }, "escalation": { "type": "array", "items": { "type": "string" } } } }, "lifecycle": { "type": "object", "properties": { "createdDate": { "type": "string" }, "plannedRemovalDate": { "type": "string" }, "cleanupTasks": { "type": "array", "items": { "type": "string" } } } } } }
javascriptconst flagPlan = await executeSkill('feature-flags', { mode: 'rollout', feature: { name: 'New Dashboard Experience', description: 'Redesigned analytics dashboard with AI insights', type: 'release', owner: 'product-team', impactLevel: 'high' }, rolloutStrategy: { type: 'gradual', stages: [ { name: 'internal', percentage: 100, duration: '3 days' }, { name: 'beta', percentage: 10, duration: '1 week' }, { name: 'early-adopters', percentage: 25, duration: '1 week' }, { name: 'general', percentage: 100, duration: 'permanent' } ] } });
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 32,821 | 23,591 | -28% | 1 | 1 | 0% | 6,506 | 5,031 | -23% | 0 | 0 | — |
case-02 | fail→pass | 38,682 | 22,036 | -43% | 1 | 1 | 0% | 6,236 | 5,462 | -12% | 0 | 0 | — |
case-03 | fail→pass | 46,175 | 30,198 | -35% | 1 | 1 | 0% | 7,250 | 5,945 | -18% | 0 | 0 | — |
case-04 | pass→pass | 13,676 | 22,491 | +64% | 1 | 1 | 0% | 2,725 | 5,219 | +92% | 0 | 0 | — |
case-05 | pass→pass | 9,562 | 10,852 | +13% | 1 | 1 | 0% | 2,011 | 3,770 | +87% | 0 | 0 | — |
case-06 | pass→pass | 10,840 | 9,490 | -12% | 1 | 1 | 0% | 1,931 | 3,300 | +71% | 0 | 0 | — |
case-07 | fail→pass | 19,386 | 13,939 | -28% | 1 | 1 | 0% | 3,324 | 4,237 | +27% | 0 | 0 | — |
case-08 | fail→pass | 15,718 | 15,253 | -3% | 1 | 1 | 0% | 2,953 | 4,215 | +43% | 0 | 0 | — |
case-09 | fail→pass | 31,959 | 14,088 | -56% | 1 | 1 | 0% | 2,160 | 4,328 | +100% | 0 | 0 | — |
case-10 | fail→pass | 16,746 | 19,393 | +16% | 1 | 1 | 0% | 3,087 | 4,466 | +45% | 0 | 0 | — |
case-11 | fail→pass | 9,818 | 12,659 | +29% | 1 | 1 | 0% | 1,623 | 4,266 | +163% | 0 | 0 | — |
case-12 | fail→pass | 12,567 | 11,116 | -12% | 1 | 1 | 0% | 2,469 | 3,553 | +44% | 0 | 0 | — |
case-13 | fail→pass | 15,709 | 20,228 | +29% | 1 | 1 | 0% | 2,809 | 4,669 | +66% | 0 | 0 | — |
case-14 | fail→pass | 11,316 | 13,723 | +21% | 1 | 1 | 0% | 1,846 | 4,002 | +117% | 0 | 0 | — |
case-15 | fail→pass | 12,694 | 13,710 | +8% | 1 | 1 | 0% | 2,571 | 4,322 | +68% | 0 | 0 | — |
case-16 | fail→pass | 15,226 | 20,627 | +35% | 1 | 1 | 0% | 2,456 | 5,698 | +132% | 0 | 0 | — |
case-17 | fail→pass | 15,116 | 11,732 | -22% | 1 | 1 | 0% | 3,232 | 3,686 | +14% | 0 | 0 | — |
case-18 | fail→pass | 10,465 | 12,045 | +15% | 1 | 1 | 0% | 1,875 | 3,634 | +94% | 0 | 0 | — |
case-19 | fail→pass | 20,447 | 14,937 | -27% | 1 | 1 | 0% | 3,312 | 4,423 | +34% | 0 | 0 | — |
case-20 | fail→pass | 13,602 | 18,370 | +35% | 1 | 1 | 0% | 2,228 | 5,243 | +135% | 0 | 0 | — |
case-21 | pass→pass | 7,589 | 9,037 | +19% | 1 | 1 | 0% | 1,554 | 3,194 | +106% | 0 | 0 | — |
case-22 | fail→pass | 14,871 | 14,907 | +0% | 1 | 1 | 0% | 3,060 | 4,227 | +38% | 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 21 counted toward the lift figure. The other 1 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 +77 percentage points is the difference between those two pass rates over the 21 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.