AI Skills

Systematically debug issues using graph-powered code navigation

Use when creating or editing skills, before deployment, to verify they work under pressure and resist rationalization - applies RED-GREEN-REFACTOR cycle to process documentation by running baseline without skill, writing to address failures, iterating to close loopholes

When the user wants to create, optimize, or audit migration guides for users switching from competitors. Also use when the user mentions "migration guide," "migrate from X," "switch to [product]," "import from X," or "data migration." For rebrand and redirects, use rebranding-strategy.

Design click/first-click tests to evaluate navigation and information findability.

收到代码审查反馈后、实施建议之前使用，尤其当反馈不明确或技术上有疑问时——需要技术严谨性和验证，而非敷衍附和或盲目执行

>

Summarize a user interview transcript into structured insights with key themes, quotes, and action items. Use after conducting user interviews to extract and share findings efficiently.

Use when handling external API rate limits (e.g., OpenAI 429s, HubSpot or Stripe rate limits), preventing duplicate work from rapid event bursts (debouncing user actions), spreading load over time, ensuring per-tenant fairness, processing events in batches, limiting concurrent runs of the same operation, or assigning priority to important runs. Covers Inngest flow control: concurrency limits with keys, throttling, rate limiting, debounce, priority, singleton, and event batching.

Deep learning framework (PyTorch Lightning / lightning package). Organize PyTorch code into LightningModules, configure Trainers for multi-GPU/TPU, implement data pipelines, callbacks, logging (W&B, TensorBoard, MLflow), distributed training (DDP, FSDP, DeepSpeed), for scalable neural network training.

|

Use when the user needs human-in-the-loop workflows in Airflow (approval/reject, form input, or human-driven branching). Covers ApprovalOperator, HITLOperator, HITLBranchOperator, HITLEntryOperator, HITLTrigger. Requires Airflow 3.1+. Does not cover AI/LLM calls (see airflow-ai).

>

Use when designing a Go constructor or factory function with optional configuration — especially with 3+ optional parameters or extensible APIs. Also use when building a New* function that takes many settings, even if they don't mention "functional options" by name. Does not cover general function design (see go-functions).

Shared backend guide for Langfuse's Next.js, tRPC, BullMQ, and TypeScript monorepo. Use when creating or reviewing tRPC routers, public REST endpoints, BullMQ queue processors, backend services, middleware, Prisma or ClickHouse data access, OpenTelemetry instrumentation, Zod validation, env configuration, or backend tests across web, worker, or packages/shared.

Identify, evaluate, and prioritize design opportunities using impact-effort frameworks and strategic criteria.

Compatibility-first Claude CLI reimplementation with faster startup, lower memory, and drop-in command compatibility

在宣称工作完成、已修复或测试通过之前使用，在提交或创建 PR 之前——必须运行验证命令并确认输出后才能声称成功；始终用证据支撑断言

>

Interactive AI-driven game for learning Docker, Linux, networking, and Kubernetes through story-driven challenges in your AI editor.

Catalog of MSBuild anti-patterns with detection rules and fix recipes. USE FOR: reviewing, auditing, or cleaning up .csproj, .vbproj, .fsproj, .props, .targets, or .proj files. Each anti-pattern has a symptom, explanation, and concrete BAD→GOOD transformation. Covers Exec-instead-of-built-in-task, unquoted conditions, hardcoded paths, restating SDK defaults, scattered package versions, and more. DO NOT USE FOR: non-MSBuild build systems (npm, Maven, CMake, etc.), project migration to SDK-style (use msbuild-modernization).

Add missing test coverage for your local code changes by generating new test files (covers uncommitted and untracked changes, or the latest commit if everything is committed). Use when you want write tests for new logic or increase test coverage.

遇到任何 bug、测试失败或异常行为时使用，在提出修复方案之前执行

When the user wants to find patterns in what content works and what doesn't. Also use when the user mentions 'what's working,' 'content patterns,' 'best topics,' 'best format,' 'best time to post,' 'analyze my content,' 'do more of,' 'do less of,' or 'what should I change.' For raw metrics, see performance-analyzer-sms. For audience-specific analysis, see audience-growth-tracker-sms. For actionable recommendations, see optimization-advisor-sms.

>

Use when working with Go slices, maps, or arrays — choosing between new and make, using append, declaring empty slices (nil vs literal for JSON), implementing sets with maps, and copying data at boundaries. Also use when building or manipulating collections, even if the user doesn't ask about allocation idioms. Does not cover concurrent data structure safety (see go-concurrency).

When the user wants to run CTV, OTT, or streaming TV ads. Also use when the user mentions "CTV ads," "connected TV," "OTT advertising," "streaming ads," "TV ads," "Hulu ads," "Roku ads," "YouTube TV ads," or "programmatic TV." For paid mix, use paid-ads-strategy.

Establish design review gates with criteria, checklists, and approval workflows.

MiniStack is a free, open-source local AWS emulator (LocalStack replacement) that emulates 25+ AWS services on a single port with no account or license required.

Quantifies market breadth health using TraderMonty's public CSV data. Generates a 0-100 composite score across 6 components (100 = healthy). No API key required. Use when user asks about market breadth, participation rate, advance-decline health, whether the rally is broad-based, or general market health assessment.

Grilling session that mines the user for fragments — heterogeneous nuggets of writing (claims, vignettes, sharp sentences, half-thoughts) — and appends them to a single document as raw material for a future article. Use when the user wants to develop ideas before imposing structure, or mentions "fragments", "ideate", or "raw material" for writing.