dethan3

Review code changes for correctness, security, performance, and style. Use when the user wants a thorough review of a diff, PR, or specific files before merging or submitting.

Systematically debug an error, unexpected behavior, or failing test. Use when the user has a bug, crash, or test failure they need to diagnose and fix.

Generate or update documentation for an existing codebase. Use when the user needs a README, API reference, module-level docs, or inline code comments written or refreshed based on the actual code.

Generate a well-formatted Git commit message based on staged changes. Use when the user wants to commit code and needs a clear, conventional commit message written for them.

Break down a feature or project requirement into a structured development plan with tasks, milestones, and implementation order. Use when the user has a goal but needs to figure out what to build and in what sequence.

Refactor code to improve structure, readability, or performance without changing external behavior. Use when the user wants to clean up or restructure existing code safely.

Generate a new project skeleton from scratch based on the user's chosen tech stack. Use when starting a new project and needing the initial directory structure, config files, and boilerplate set up correctly.

Generate unit tests or integration tests for existing code. Use when the user wants test coverage for a function, module, or feature that currently has no tests or insufficient tests.

像松鼠 Nutty 一样收藏和整理知识的坚果小助手。自动保存链接、想法、项目、任务到飞书多维表格。使用场景：发送URL、记录想法、创建项目、添加任务、添加标签、提及相关项目，或查询已保存的个人知识。优先保存而非分析，保留原始输入，路由到链接/想法/项目/任务/收件箱，并简短回复。

检查中文文本中的错别字和语病，支持代码注释、文档、纯文本。

Summarize raw subtitle and transcript files such as `.srt`, `.vtt`, `.txt`, and interview transcripts into structured briefing documents. Use when an agent needs to process long-form podcasts, interviews, technical talks, or panel discussions stored in `raw/`, especially when timestamps, fragmented captions, multi-speaker debate, hard numbers, and actionable insights must be reconstructed into a coherent summary.

Build and maintain a persistent Feishu/Lark knowledge base using the LLM Wiki pattern. Use when the user wants to ingest new documents into an interlinked wiki, query accumulated knowledge, or lint the wiki for gaps, contradictions, and stale pages. Best when `lark-cli` is installed and a `wiki-instance.yaml` config has been filled in.

Broad AI ecosystem scout. Use when an agent needs to map, rank, or analyze promising AI projects, applications, platforms, model/tool ecosystems, connectors, agent products, model-serving stacks, fine-tuning tools, eval/observability products, workflow builders, memory/context systems, and contribution opportunities across the full AI landscape.

GitHub-first open source contribution assistant for coding agents. Use when you need to inspect origin/upstream/fork relationships, scout current issues and PRs, verify whether an issue is already claimed or overlaps active work, decide whether a problem belongs in this repository or an external API or platform, plan a fix, or draft branch, commit, and PR materials without assuming the user wants to ship yet.

Build and maintain a local markdown knowledge base from raw source materials by ingesting files, scaffolding source notes, synthesizing concept pages, rebuilding derived indexes, and running integrity checks. Use when Codex needs to turn a folder of research notes, articles, papers, exports, or mixed markdown/text documents into a structured wiki-like knowledge base with `raw/`, `sources/`, `concepts/`, `indexes/`, `outputs/`, and `state/`.

网页剪藏工具，智能抓取网页内容并保存为 Markdown 格式，支持 Obsidian 集成和知识图谱构建。 **当以下情况时使用此 Skill**： (1) 用户想要保存网页内容 (2) 用户提到"剪藏"、"保存网页"、"收藏文章" (3) 用户提供 URL 并希望保存内容 (4) 用户需要将网页内容整理到飞书文档或本地文件 (5) 用户需要批量处理多个网页链接 **激活关键词**： - 剪藏、保存网页、收藏文章、网页存档 - claw clip、web clipper、save webpage - 保存到飞书、保存到 Obsidian - 批量保存、批量处理链接