发现 Skills

当需要评估产品的隐私合规状态或生成合规评估报告时使用。隐私合规评估文档自动生成，覆盖个人信息收集、存储、使用、共享全生命周期，对标GDPR/PIPL等法规，输出结构化Markdown评估文档。关键词：隐私合规、GDPR、PIPL、个人信息保护、数据合规、隐私评估、合规评估。

当需要深入理解需求背后的真实问题时使用。需求自动理解与拆解，将分类后的需求还原为真实问题，分析需求间的关联关系。执行角色：🤖→👤 AI建议，人类审批。关键词：需求理解、需求拆解、问题还原、需求关联、需求分析。

Design robust tool-using AI agent workflows. Use when Codex needs to specify an agent's goal loop, tool boundaries, state model, failure handling, approval gates, verification path, or handoff contract for multi-step agent systems, coding agents, research agents, PM agents, or productized automations.

How to configure MCP servers in .rkat/mcp.toml

How to use task_create/task_update/task_list for structured work tracking

Build, debug, and optimize Claude API / Anthropic SDK apps. Apps built with this skill should include prompt caching. TRIGGER when: code imports anthropic/@anthropic-ai/sdk; user asks to use the Claude API, Anthropic SDKs, or Managed Agents (/v1/agents, /v1/sessions, /v1/environments). DO NOT TRIGGER when: code imports `openai`/other AI SDK, general programming, or ML/data-science tasks.

Pre-submission audit of iOS/macOS apps against 150+ Apple App Store Review Guidelines. Analyzes source code, project config, metadata, and screenshots to catch rejection risks before you submit. Provides per-guideline verdicts (PASS/FAIL/WARNING/N/A) with specific fix suggestions. Use this skill whenever someone wants to check if their app will pass App Store review, asks about App Store rejection risks, says "will Apple approve this", "check my app for review", "pre-submission audit", "App Store compliance check", "why might my app get rejected", "review guidelines check", or mentions preparing an app for App Store submission — even if they don't say "review guidelines" explicitly. Also trigger when someone is debugging a rejection and wants to know what else might fail.

Version bump, changelog finalize, build, test, tag, publish to registries, and create GitHub Release. Language-aware for TypeScript (npm), Python (PyPI), and Rust (crates.io). Delegates to manage-git for tagging strategy and manage-documentation for changelog format. Use when ready to release a new version.

Comprehensive repository analysis using Explore agents, web search, and Context7 to investigate codebase structure, technology stack, configuration, documentation quality, and provide actionable insights. Use this skill when asked to analyze, audit, investigate, or report on a repository or codebase.

Tests software functionality and validates user flows through the browser. Runs unit tests, integration tests, and E2E browser automation. Generates a structured QA report with pass/fail results and screenshots.

A professional Sudoku coach that generates valid puzzles (3×3 to 6×6 grid sizes) with adjustable difficulty (1–5 stars) and provides printable card prompts.

Search and retrieve academic papers from arXiv using their free REST API. No API key needed. Search by keyword, author, category, or ID. Combine with web_extract or the ocr-and-documents skill to read full paper content.

Production pipeline for mathematical and technical animations using Manim Community Edition. Creates 3Blue1Brown-style explainer videos, algorithm visualizations, equation derivations, architecture diagrams, and data stories. Use when users request: animated explanations, math animations, concept visualizations, algorithm walkthroughs, technical explainers, 3Blue1Brown style videos, or any programmatic animation with geometric/mathematical content.

Synthesizes intrinsic (DCF) and relative (multiples) valuation outputs into a final value estimate and investment recommendation. Reconciles divergent valuations, reverse-engineers what the market is pricing in (implied growth, implied ROIC), computes margin of safety, and produces a buy/sell/hold recommendation with catalysts. Use when combining multiple valuation approaches, making investment recommendations, reconciling DCF with multiples, or when user mentions reconcile valuations, investment recommendation, margin of safety, implied growth rate, or what is the market pricing in.

Systematic approach to analyzing compiled binaries, understanding program behavior, and identifying vulnerabilities without source code access

Use this skill to diagnose Agent Skills execution, daemon, lifecycle, readIndex, timeout, and log problems.

Use this starter when a task lands in the `Differentiable simulation` frontier leaf and the repository has curated resources but no dedicated runtime implementation yet.

DigitalOcean Managed Databases for PostgreSQL, MySQL, Redis, MongoDB, Kafka, OpenSearch, and Valkey. Use when provisioning, scaling, or operating managed database clusters on DigitalOcean.

Quality gate. 5 parallel agents review changes. All must pass.

Build a compact, high-signal context brief (files, symbols, recent commits) instead of pasting large code blocks

Execute tasks from track plan with TDD workflow and git commit integration

Task lifecycle state transitions with validation gates. Defines states, triggers, and required proofs.

AI-assisted video editing workflows for cutting, structuring, and augmenting real footage. Covers the full pipeline from raw capture through FFmpeg, Remotion, ElevenLabs, fal.ai, and final polish in Descript or CapCut. Use when the user wants to edit video, cut footage, create vlogs, or build video content.

Generate or update project documentation from code -- API docs, service docs, architecture docs, or runbooks. Use after adding endpoints, entities, events, or infrastructure changes, or when documentation may be stale.