feature-flags
LaunchDarkly/Flagsmith/Vercel patterns, gradual rollouts, percentage splits, A/B toggles.
Works with
---
name: feature-flags
description: LaunchDarkly/Flagsmith/Vercel patterns, gradual rollouts, percentage splits, A/B toggles.
license: MIT
---
# Feature Flags
## Rules
- Use feature flags for gradual rollouts — ship code behind flags, enable incrementally
- Flag types: boolean (on/off), percentage (10% of users), user-targeting (specific IDs/segments)
- LaunchDarkly for enterprise, Flagsmith for open-source, Vercel Flags or `@vercel/flags` for Next.js
- Evaluate flags server-side by default — only pass flag values to client, never the evaluation logic
- Flag naming convention: `enable-new-checkout`, `show-pricing-v2` — descriptive, kebab-case
- Clean up flags after full rollout — stale flags are tech debt; set expiration dates
- Percentage rollouts: hash user ID for consistent bucketing — same user always sees same variant
- Default to flag-off for new features — fail safe if flag service is unavailable
- Use flags for A/B testing: define variants, track conversion events, measure statistical significance
- Store flag overrides in environment variables for local development
## Patterns
```ts
// Vercel Flags / Next.js pattern
import { flag } from "@vercel/flags/next";
export const showNewCheckout = flag({
key: "show-new-checkout",
decide: () => false, // default off
});
// In server component
const enabled = await showNewCheckout();
```
```ts
// LaunchDarkly pattern
import * as ld from "@launchdarkly/node-server-sdk";
const client = ld.init(process.env.LAUNCHDARKLY_SDK_KEY!);
const showFeature = await client.variation("new-checkout", user, false);
```
## Avoid
- Shipping features without a flag — every new feature should be toggleable
- Nested flag dependencies — keep flag logic flat and independent
- Leaving flags in code after 100% rollout — schedule cleanup within 2 weeks
- Client-side flag evaluation exposing targeting rules — evaluate on server, send resultMore Deployment & CI/CD skills
finetuning
microsoft/azure-skills
Fine-tune models on Microsoft Foundry using SFT (supervised), DPO (preference), or RFT (reinforcement with graders). Covers dataset preparation, training job submission, deployment, and evaluation. USE FOR: fine-tune, SFT, DPO, RFT, training data, grader, distillation, fine-tuned model, training job, large file upload, calibrate grader, deploy fine-tuned model, evaluate fine-tuned model. DO NOT USE FOR: general model deployment without fine-tuning (use deploy-model), agent creation (use agents), prompt optimization without training (use prompt-optimizer).
prisma-compute
prisma/skills
Prisma Compute deployment and hosting guide. Use whenever the user mentions Prisma Compute, `prisma.compute.ts`, `defineComputeConfig`, deploying or hosting a Prisma app, `@prisma/cli app deploy`, `compute:deploy`, `create-prisma --deploy`, `PRISMA_SERVICE_TOKEN`, Compute auth/workspaces, apps/deployments/build logs/domains, localhost vs `0.0.0.0`, deploy port binding, or framework deploy readiness for Hono, Elysia, Next.js, TanStack Start, Astro, Nuxt, Svelte, Nest, Turborepo, or custom/prebuilt artifacts.
azure-quotas
microsoft/azure-skills
Check/manage Azure quotas and usage across providers. For deployment planning, capacity validation, region selection. WHEN: \"check quotas\", \"service limits\", \"current usage\", \"request quota increase\", \"quota exceeded\", \"validate capacity\", \"regional availability\", \"provisioning limits\", \"vCPU limit\", \"how many vCPUs available in my subscription\".

