ai-infrastructure-hugg…
Hugging Face Inference SDK patterns for TypeScript/Node.js — InferenceClient setup, chat completion, text generation, streaming, embeddings, image generation,…
SST (Ion) infrastructure-as-code — TypeScript-first serverless on AWS with Pulumi, resource linking, and live Lambda dev
$ npx -y skills add agents-inc/skills --skill infra-iac-sst --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/infra-iac-sstContext preview
The summary Claude sees to decide when to auto-load this skill.
SST (Ion) infrastructure-as-code — TypeScript-first serverless on AWS with Pulumi, resource linking, and live Lambda dev
name: infra-iac-sst description: SST (Ion) infrastructure-as-code — TypeScript-first serverless on AWS with Pulumi, resource linking, and live Lambda dev
> **Quick Guide:** SST v3 (Ion) is TypeScript-first infrastructure-as-code for AWS, powered by Pulumi/Terraform (not CDK/CloudFormation). Define your entire app in `sst.config.ts` using high-level components (`sst.aws.Function`, `sst.aws.ApiGatewayV2`, `sst.aws.Bucket`, `sst.aws.Dynamo`, etc.). Use **resource linking** (`link: [bucket]` + `Resource.MyBucket.name`) for type-safe, permission-aware access between components. Use `sst dev` for live Lambda development with sub-10ms reloads. Use `$app.stage` for multi-environment isolation. Use `transform` to customize underlying Pulumi resources.
---
<critical_requirements>
> **All code must follow project conventions in CLAUDE.md** (kebab-case, named exports, import ordering, `import type`, named constants)
**(You MUST use resource linking (`link` + `Resource.*`) to connect components — NEVER hardcode ARNs, table names, or bucket names)**
**(You MUST use `$app.stage` for environment isolation — NEVER share resources across stages without explicit intent)**
**(You MUST use `sst dev` for local development — it provides live Lambda proxying with sub-10ms reloads against real AWS resources)**
**(You MUST use `sst secret set` for secrets — NEVER put secrets in `sst.config.ts`, `.env` files committed to git, or environment variables)**
**(You MUST use `transform` to customize underlying resources — NEVER reach for raw Pulumi resources when an SST component exists)**
</critical_requirements>
---
---
**Auto-detection:** SST, sst.config.ts, sst.aws.Function, sst.aws.ApiGatewayV2, sst.aws.Bucket, sst.aws.Dynamo, sst.aws.Queue, sst.aws.Topic, sst.aws.Cron, sst.aws.Nextjs, sst.aws.Remix, sst.aws.Astro, sst.aws.StaticSite, sst.aws.Vpc, sst.aws.Cluster, sst.aws.Postgres, sst.aws.Router, sst.Linkable, Resource from sst, sst dev, sst deploy, sst remove, sst secret, $app.stage, $transform, $concat, $interpolate, resource linking, live Lambda, Ion
**When to use:**
**When NOT to use:**
**Key patterns covered:**
---
<philosophy>
SST v3 (Ion) replaces CDK/CloudFormation with Pulumi/Terraform for dramatically faster deployments and a simpler programming model. The core ideas:
1. **One config file** — Your entire app is defined in `sst.config.ts`. Infrastructure, frontends, and functions all declared together in TypeScript with loops, conditionals, and functions. 2. **Components over constructs** — High-level `sst.aws.*` components encapsulate best practices (IAM, logging, monitoring). Use `transform` to reach into underlying resources when defaults aren't enough. 3. **Resource linking** — The killer feature. `link: [bucket]` automatically grants IAM permissions and injects type-safe references. Access via `Resource.MyBucket.name` at runtime. No manual ARN passing or environment variable wiring. 4. **Stage-based isolation** — Every developer gets their own stage (`sst dev` creates a personal stack). `$app.stage` drives resource naming. Production uses `sst deploy --stage production`. 5. **Live dev against real AWS** — `sst dev` replaces Lambda functions with stubs that proxy to your local machine. Changes reload in under 10ms. No local emulation — your code runs against real DynamoDB, S3, SQS.
**When to use SST:**
**When NOT to use SST:**
</philosophy>
---
<patterns>
Every SST app has a single `sst.config.ts` with two functions: `app()` for metadata and `run()` for resou
The official skills marketplace for Agents Inc. 150+ skills covering everything from React and Prisma to Redis, ElevenLabs, and infrastructure tooling. Pick the skills that match your stack and install them via Claude Code. Need more control?
Repo: agents-inc/skills
Hugging Face Inference SDK patterns for TypeScript/Node.js — InferenceClient setup, chat completion, text generation, streaming, embeddings, image generation,…
LiteLLM proxy server setup, TypeScript client patterns via OpenAI SDK, model routing, fallbacks, load balancing, spend tracking, virtual keys, and production…
Serverless GPU compute platform for AI model deployment — web endpoints, GPU functions, model serving, and TypeScript client patterns
Local LLM inference with the Ollama JavaScript client -- chat, streaming, tool calling, vision, embeddings, structured output, model management, and…
Replicate SDK patterns for TypeScript/Node.js -- client setup, predictions, streaming, webhooks, file handling, model versioning, deployments, and training
Together AI SDK patterns for TypeScript — client setup, chat completions, streaming, structured output, function calling, embeddings, image generation,…