Best Claude Code skills for ai & agent development on Api (August 2026)

Reference for the Claude API / Anthropic SDK — model ids, pricing, params, streaming, tool use, MCP, agents, caching, token counting, model migration. TRIGGER — read BEFORE opening the target file; don't skip because it "looks like a one-liner" — whenever: the prompt names Claud…

Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).

/cs:caio-review <plan> — Eval-demanding Chief AI Officer interrogation of any plan that involves AI: model selection, risk classification, cost economics, or AI hiring. Use when shipping an AI feature without an eval set, choosing between API, fine-tune, and self-hosted, or clas…

/cs:cfo-review <plan> — Numerate-skeptic interrogation of any plan that touches money. Unit economics, runway, dilution, capital allocation. Use when a plan commits meaningful spend — e.g. a hiring wave, a fundraise decision, or a new channel budget.

Chief AI Officer advisory for startups: model build-vs-buy decisions (API vs fine-tune vs in-house), AI risk classification under EU AI Act + US state patchwork, AI cost economics (API-to-self-hosted breakeven), and AI team org evolution. Use when deciding whether to call an API…

ML engineering skill for productionizing models, building MLOps pipelines, and integrating LLMs. Covers model deployment, feature stores, drift monitoring, RAG systems, and cost optimization. Use when the user asks about deploying ML models to production, setting up MLOps infras…

Generates unit tests, integration tests, and E2E tests for React/Next.js applications. Scans components to create Jest + React Testing Library test stubs, analyzes Istanbul/LCOV coverage reports to surface gaps, scaffolds Playwright test files from Next.js routes, mocks API call…

Use proactively whenever LLM API costs come up -- or should. Triggers include: 'my AI costs are too high', 'optimize token usage', 'which model should I use', 'LLM spend is out of control', 'implement prompt caching', 'we're about to launch an AI feature', 'build me an AI endpoi…

Reverse-engineer any codebase into a complete Product Requirements Document (PRD). Analyzes routes, components, state management, API integrations, and user interactions to produce business-readable documentation detailed enough for engineers or AI agents to fully reconstruct ev…

Generates complete, production-ready SaaS project boilerplate including authentication, database schemas, billing integration, API routes, and a working dashboard using Next.js 14+ App Router, TypeScript, Tailwind CSS, shadcn/ui, Drizzle ORM, and Stripe. Use when the user wants …

Use when the user says 'build me an app', 'create a project from this spec', 'scaffold a new repo', 'generate a starter', 'turn this idea into code', 'bootstrap a project', 'I have requirements and need a codebase', or provides a natural-language project specification and expect…

Use when managing the money for an internal R&D program or portfolio — building a multi-period program budget with the F&A (indirect) split, tracking burn rate and runway against value-inflection milestones, or routing R&D cost items to a capitalize-vs-expense determination. Eve…

Use when planning, funding, scoping, or synthesizing enterprise research across workstreams — clinical study design, R&D program finance, market sizing/surveys, or product/user research. Triggers on "design this clinical study", "what sample size", "R&D budget", "burn rate", "ca…

Decision-grade entity research skill — produces a hypothesis-tested dossier on a specific company, person, nonprofit, or government org, not a generic profile. Forcing intake makes the user state their hypothesis upfront (what they already believe and want to verify or disprove)…

Vercel AI SDK v5 for backend AI (text generation, structured output, tools, agents). Multi-provider. Use for server-side AI or encountering AI_APICallError, AI_NoObjectGeneratedError, streaming failures.

Master REST and GraphQL API design principles to build intuitive, scalable, and maintainable APIs that delight developers. Use when designing new APIs, reviewing API specifications, or establishing API design standards.

Implements API rate limiting using token bucket, sliding window, and Redis-based algorithms to protect against abuse. Use when securing public APIs, implementing tiered access, or preventing denial-of-service attacks.

Comprehensive Cloudflare account management for deploying Workers, KV Storage, R2, Pages, DNS, and Routes. Use when deploying cloudflare services, managing worker containers, configuring KV/R2 storage, or setting up DNS/routing. Requires CLOUDFLARE_API_KEY in .env and Bun runtim…

Cloudflare Workers security with authentication, CORS, rate limiting, input validation. Use for securing APIs, JWT/API keys, or encountering auth failures, CORS errors, XSS/injection vulnerabilities.

| Type-safe ORM for Cloudflare D1 databases using Drizzle. Use when: building D1 database schemas, writing type-safe SQL queries, managing migrations with Drizzle Kit, defining table relations, implementing prepared statements, using D1 batch API, or encountering D1_ERROR, trans…

Firecrawl v2.5 API for web scraping/crawling to LLM-ready markdown. Use for site extraction, dynamic content, or encountering JavaScript rendering, bot detection, content loading errors.

Google Gemini API with @google/genai SDK. Use for multimodal AI, thinking mode, function calling, or encountering SDK deprecation warnings, context errors, multimodal format errors.

Google Gemini embeddings API (gemini-embedding-001) for RAG and semantic search. Use for vector search, Vectorize integration, or encountering dimension mismatches, rate limits, text truncation.

Idempotent API operations with idempotency keys, Redis caching, DB constraints. Use for payment systems, webhook retries, safe retries, or encountering duplicate processing, race conditions, key expiry errors.

Deploy ML models with FastAPI, Docker, Kubernetes. Use for serving predictions, containerization, monitoring, drift detection, or encountering latency issues, health check failures, version conflicts.

Complete guide for OpenAI APIs: Chat Completions (GPT-5.2, GPT-4o), Embeddings, Images (GPT-Image-1.5), Audio (Whisper + TTS + Transcribe), Moderation. Includes Node.js SDK and fetch approaches.

OpenAI Assistants API v2 for stateful chatbots with Code Interpreter, File Search, RAG. Use for threads, vector stores, or encountering active run errors, indexing delays. ⚠️ Sunset August 26, 2026.

OpenAI Responses API for stateful agentic applications with reasoning preservation. Use for MCP integration, built-in tools, background processing, or migrating from Chat Completions.

| This skill should be used when writing or reviewing Swift code for iOS or macOS projects. Apply modern Swift 6+ best practices, concurrency patterns, API design guidelines, and migration strategies. Covers async/await, actors, MainActor, Sendable, typed throws, and Swift 6 bre…

TanStack AI (alpha) provider-agnostic type-safe chat with streaming for OpenAI, Anthropic, Gemini, Ollama. Use for chat APIs, React/Solid frontends with useChat/ChatClient, isomorphic tools, tool approval flows, agent loops, multimodal inputs, or troubleshooting streaming and to…

Creates detailed technical specifications for software projects covering requirements, architecture, APIs, and testing strategies. Use when planning features, documenting system design, or creating architecture decision records.

Vercel KV (Redis-compatible key-value storage via Upstash). Use for Next.js caching, sessions, rate limiting, TTL data storage, or encountering KV_REST_API_URL errors, rate limit issues, JSON serialization errors. Provides strong consistency vs eventual consistency.

Use whenever answering anything version-sensitive — library/framework/SDK APIs, package versions, model IDs, pricing, CLI flags, config, or "latest/newest" anything. Verify against current sources instead of training data.