數據與 AI

Use when creating or modifying Claude Code slash commands - command development with frontmatter, arguments handling, and namespacing for November 2025 specifications

Guidance for 50+ specialized Wolf agent roles with responsibilities and collaboration patterns

Create new Claude Code skills with proper SKILL.md structure, frontmatter, and best practices. Use when building new skills for this repository.

Automated video translation tool. Use when the user wants to translate English videos to Chinese (subtitles and/or dubbed audio). Supports generating SRT subtitles, AI voiceover dubbing, and merging them into the final video.

shadcn/ui component library for React with Tailwind CSS - copy-paste accessible components with full code ownership. Use when building React applications needing customizable UI components with Next.js, Vite, or Remix integration.

This skill should be used when users want to train or fine-tune language models using TRL (Transformer Reinforcement Learning) on Hugging Face Jobs infrastructure. Covers SFT, DPO, GRPO and reward modeling training methods, plus GGUF conversion for local deployment. Includes guidance on the TRL Jobs package, UV scripts with PEP 723 format, dataset preparation and validation, hardware selection, cost estimation, Trackio monitoring, Hub authentication, and model persistence. Should be invoked for tasks involving cloud GPU training, GGUF conversion, or when users mention training on Hugging Face Jobs without local GPU setup.

Perform in-depth audits of Rust code, covering ownership, concurrency, unsafe blocks, traits, and Cargo dependencies.

Scala 3.6+ best practices with ScalaTest 3.2, sbt 1.10, functional programming patterns, and Play Framework.

External API integrations (Perplexity AI, Gemini, Binance, CoinGecko). TRIGGERS: 'API', 'Perplexity', 'Gemini', 'Binance', 'CoinGecko', 'rate limit', 'API error'. Use when implementing API connections, handling rate limits, error recovery, and data transformation patterns.

Node.js backend development with TypeScript, Express/Fastify servers, routing, middleware, and database integration

ElevenLabs Speech-to-Text transcription workflows with Scribe v1 supporting 99 languages, speaker diarization, and Vercel AI SDK integration. Use when implementing audio transcription, building STT features, integrating speech-to-text, setting up Vercel AI SDK with ElevenLabs, or when user mentions transcription, STT, Scribe v1, audio-to-text, speaker diarization, or multi-language transcription.

MANDATORY for any web UI or styling change. Must be followed whenever touching web/src UI, layout, or CSS. Absolutely no shadows, glows, or flares; keep everything flat and token-driven.

Check currently firing Grafana alerts, analyze alert status, and investigate alert issues in the Kagenti platform

Use this skill when styling pages, creating components, working with CSS variables, design patterns, or applying the Token Milagre visual identity. Contains spacing rules, color system, card effects, and component patterns.

Project Memory Management adapted from Context Engineering Kit. Enables the agent to maintain a persistent ".opencode/MEMORY.md" file to store architectural decisions, patterns, and "lessons learned".

Enterprise multi-agent workflow orchestration specialist. Master workflow design, agent coordination, task delegation, and process automation with Context7 MCP integration and comprehensive monitoring. Build scalable, intelligent workflow systems with fault tolerance and performance optimization.

Systematically validate implementation plans before execution - catches hallucinations, TDD violations, missing references, and architectural issues that derail implementations. Use after writing-plans skill creates a plan document.

Write secure, performant database queries using parameterized queries, proper eager loading, and transaction management. Use this skill when writing database queries, ORM operations, raw SQL statements, or repository methods. When working on files in repositories/, queries/, services/, or any code that interacts with the database. When optimizing query performance, preventing N+1 queries, implementing transactions, adding query timeouts, or caching expensive queries. When reviewing code for SQL injection vulnerabilities.

Analyze and optimize Django ORM queries including N+1 problems, missing indexes, slow queries, and migration issues. Use when troubleshooting slow API responses, database performance, query optimization, or migration errors.

Generate new domain entities with full CRUD infrastructure in noop-based projects. Use when adding new resources, tables, or API endpoints to an existing project.