Code quality assurance strategies. Parent skill for quality-related skills: security (security review) and debug (systematic debugging).

以资深架构师视角对 Diff/源码做深度 Code Review：正确性与防御性、架构合规、DRY、KISS、测试完备性；禁止格式挑刺；单次最多 5 条高价值问题；无重大问题输出 LGTM。在用户附上改动、请求审查、PR review、走查、diff 评审时触发。

Performs a structured engineering code review focusing on design quality, coupling, performance (especially mobile), and maintainability. Use when the user asks to review code, 检查代码, 审阅代码, 代码评估, or analyze code design/quality.

bias-assessor

对 Python 改动做“问题优先”的代码审查。适用于审查 diff、定位行为回归、评估风险、检查测试覆盖缺口，尤其是 crawlers、storage、fish_intel_mvp 相关改动。

合同审核技能，通过添加批注方式审查合同（不修改原文）。采用四层审核模型（主体核验、基础审核、商务条款、法律条款），生成结构化批注（问题类型、风险原因、修订建议），风险等级通过审核人名称编码，并生成合同概要、综合审核意见和Mermaid业务流程图（含渲染图片）。输出语言遵循合同语言。支持双模式企业核验：QCC CLI 用于主体信息核验（终端直连），QCC MCP 用于风险穿透预警（深度分析）。

对广告素材进行四阶段流水线审核：定性分类 → 通用红线 → 行业专项 → 生成报告。

当用户需要对合同进行中文审核、条款修订、批注说明、模板比对、风险识别，且最终要直接输出写有逐条批注和修订痕迹的 docx 文件时，使用这个 skill。典型触发包括：审核合同、修改合同、按模板优化合同、输出修订版、输出批注版、输出修订批注版、检查定义词、检查条款逻辑、检查格式、输出清洁版。**特别注意：默认交付物必须是直接可打开的 docx 逐条批注修订版，不先向用户阶段性输出 md 或 txt 征求确认；凡涉及修订和批注，一律通过直接操作 DOCX 底层 XML 写入，不允许切换为其他实现路径，确保 WPS 兼容性。**

这是一个专门用于将GPU代码适配到华为昇腾NPU的Agent Skill。本技能覆盖完整的适配工作流：代码分析、堵点识别、适配脚本编写、验证方案设计、以及最终报告生成。

# Review Writing — 学术综述逐节写作方法论

SlowMist AI Agent Security Review — comprehensive security framework for skills, repositories, URLs, on-chain addresses, and products (Claude Code version)

AI/ML 人才搜索、论文作者发现、实验室成员爬取、GitHub 研究者挖掘与个性化招聘邮件生成 skill。只要用户提到查找 AI/ML PhD、研究员、工程师，抓取实验室成员、OpenReview/CVF 会议作者、GitHub 网络研究者，提取主页/Scholar/GitHub/邮箱/研究方向，识别华人、分类去重，或把结果导入飞书多维表格并批量生成邮件，就应该优先使用这个 skill；即使用户没有明确说“使用 Mapping-Skill”，只要任务属于这些复合工作流，也应触发。

Everything Claude Code (ECC) system integration - 30 agents, 77 rules, 12 languages, 4 iron laws, TDD workflow, verification gates

绩效面谈报告、OKR对齐度检测、校准辅助

自动化 GitLab MR 代码审核流水线。使用 AI 对 MR 进行代码审查，生成报告并邮件发送给提交人。

Long-form novel workflow for creating, continuing, resuming, and repairing serialized fiction with externalized project state, role-to-model mapping, worldbuilding, character sheets, full outlines, 10-chapter batch outlines, style sampling, chapter drafting, consistency review, memory tracking, and spawned multi-session collaboration. Use when the user asks to start a novel project, continue or resume a draft, recover from truncation, assign models to roles, generate canon or chapters, review for consistency, or maintain a long-running fiction project across many chapters. Supports single-agent or multi-agent execution, with multi-agent as the default; when multi-agent is selected, first surface the available model inventory and the novel-writing role list, then ask the user for an explicit role→model mapping before any canon work. Once the user has provided the mapping, persist it in project state and drive stage work with `sessions_spawn` using the mapped roles rather than treating the mapping as passive metadata. The main session may only create the project shell and route work; it must not author canon files.

Security guard for OpenClaw skills, developed and maintained by rose北港（小红帽 / 猫猫帽帽）. Audits installed or incoming skills with local rules plus Zenmux AI intent review, then recommends pass, warn, block, or quarantine.

淘宝/天猫商品评价分析，提取用户真实反馈，发现质量问题和卖点。适合电商卖家、选品分析、市场调研。

Generate Xiaohongshu (小红书/RED) content optimized for the platform's CES algorithm. Use when: (1) creating xiaohongshu/小红书 posts, (2) writing Chinese social media content for RED, (3) generating content with xiaohongshu SEO optimization, (4) planning xiaohongshu content calendars. Supports diary-style, tutorial, review, and list formats with proper AI content labeling.

代码审查能力，调用CodeReviewer专家进行安全、性能、风格全面审查，输出问题列表和改进建议

>-

Drives the BillClaw local bookkeeping CLI against SQLite (db/expenses.db) via scripts/main.py JSON subcommands—add/query transactions, delete and category-merge with preview+confirm, user-defined categories, reports with chart PNGs, CSV export, and a local Flask dashboard. Use when the user tracks 记账/收支/账本, manages 分类, asks for 报表 or charts, wants to open the Web 看板, exports CSV, or mentions BillClaw or running main.py (root shim) / scripts/main.py.

梳理某个 Skill、脚本或工作流需要的权限，并提出最小权限替代方案。；use for permissions, least-privilege, security workflows；do not use for 绕过系统安全控制, 生成提权方法.

从评论、评价和反馈中提炼卖点、痛点、反对意见与应删除的话术。；use for reviews, voice-of-customer, marketing workflows；do not use for 造假好评, 泄露用户身份.

从剪贴板读取审稿意见、导师批注或评审反馈，生成逐条回复、修改计划与优先级建议。

在执行 shell 方案前检查危险模式，如 pipe-to-shell、覆盖式删除、危险重定向或混淆执行。；use for shell, security, command-review workflows；do not use for 提供攻击性命令, 帮用户绕过限制.

Convert a syllabus, exam scope, or course notes into a revision calendar

Turn scattered notes, metrics, and unfinished tasks into a weekly operating

汇总本周完成、阻塞、经验教训与下周计划，形成管理层版和执行版摘要。；use for weekly-review, planning, retrospective workflows；do not use for 杜撰完成项, 替代工时系统.

vwu.ai 平台上的 gemini 模型调用技能。

vwu.ai 平台上的 veo 模型调用技能。

Upload any contract or legal document and get a structured risk analysis with flagged clauses, plain-language explanations, and negotiation suggestions.

AI短剧内容基因检测。零依赖可用：AI直接分析版权疑似、年龄分级违规、小说魔改程度，输出结构化风险图谱。可选配 Python 环境启用三重相似度算法（n-gram / 编辑距离 / TF-IDF）提升精度。结果仅供内容审核参考，不构成法律意见。

门诊慢病审核（糖尿病/高血压）。输入 OCR 结果数组 JSON，输出审核结论与原因（原始 JSON + 自然语言结论）。

重大疾病理赔智能评估（支持 28 种病种）。输入住院病历结构化数据，调用内网评估接口，输出原始 JSON 与自然语言结论（结论 + 证据）。

This skill should be used when working with memory-lancedb-pro, a production-grade long-term memory MCP plugin for OpenClaw AI agents. Use when installing, configuring, or using any feature of memory-lancedb-pro including Smart Extraction, hybrid retrieval, memory lifecycle management, multi-scope isolation, self-improvement governance, or any MCP memory tools (memory_recall, memory_store, memory_forget, memory_update, memory_stats, memory_list, self_improvement_log, self_improvement_extract_skill, self_improvement_review).

Three-tier code quality defense: L1 quick scan, L2 deep audit (via bug-audit), L3 cross-validation with adversarial testing. 三级代码质量防线。

Expand a structured brief in `content-production/inbox/` into a reusable long-form markdown article draft, then run a local writer / critic / judge quality loop with a constrained humanization pass. Use when Codex needs a stage-1 article draft plus reusable writing sidecars for downstream `generate-image` and `wechat-formatter`.

Sync a local `wechat-report` result into Feishu Bitable after the user has reviewed the report and confirmed the sync.