- 因本 skill 设计缺陷导致的 bug，先用 `bensz-collect-bugs` 规范记录到 `~/.bensz-skills/bugs/`，不要直接修改用户本地已安装的 skill 源码；若有 workaround，先记 bug，再继续完成任务。

- 因本 skill 设计缺陷导致的 bug，先用 `bensz-collect-bugs` 规范记录到 `~/.bensz-skills/bugs/`，不要直接修改用户本地已安装的 skill 源码；若有 workaround，先记 bug，再继续完成任务。

上下文优化专家。专注于长对话中的上下文管理、token 效率和性能优化。解决 lost-in-middle、context poisoning 等问题，提升 AI 代理在复杂任务中的表现。

DevOps 与运维专家。精通 CI/CD、容器化、编排、基础设施即代码、监控告警和自动化部署。用于构建高效、可靠的软件交付流水线和运维系统。

文档专家。专注于技术文档编写、API 文档生成、README 优化和文档维护。提供清晰的文档结构、规范的格式和用户友好的内容。

前端开发专家。精通 React/Vue/Next.js/Nuxt 等现代前端框架，专注于视觉系统、组件化开发、响应式设计、性能优化和用户体验。用于前端应用开发、UI 设计落地、设计系统实现和高质量界面交付。

Git 工作流专家。规范化版本控制，确保提交历史清晰可追溯。支持 Conventional Commits 规范、Pull Request 最佳实践、分支管理策略和自动化工作流。

- 因本 skill 设计缺陷导致的 bug，先用 `bensz-collect-bugs` 规范记录到 `~/.bensz-skills/bugs/`，不要直接修改用户本地已安装的 skill 源码；若有 workaround，先记 bug，再继续完成任务。

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes - four-phase framework (root cause investigation, pattern analysis, hypothesis testing, implementation). NO FIXES WITHOUT ROOT CAUSE INVESTIGATION FIRST.

当用户明确要求“压缩/瘦身/精简某个 Agent Skill 的 Markdown 文档”“在不改变功能前提下降低 skill 上下文开销”时使用。先理解目标 skill 的真实能力与安全边界，再在忽略 `tests/`、`plans/` 以及目标 skill 的 `README.md`、`CHANGELOG.md` 的前提下，压缩 `SKILL.md`、`references/*.md` 等工作型 Markdown，并把中间产物隔离到 `.compact-bensz-skills/`。⚠️ 不适用：用户主要想新增功能、修复脚本逻辑、批量改代码、或只想压缩非 skill 文档。

- 因本 skill 设计缺陷导致的 bug，先用 `bensz-collect-bugs` 规范记录到 `~/.bensz-skills/bugs/`，不要直接修改用户本地已安装的 skill 源码；若有 workaround，先记 bug，再继续完成任务。

当用户明确要求“review 某个 GitHub PR”“评估某个 pull request 是否值得 merge”“帮我判断这个 PR 怎么处理”时使用。基于用户提供的 GitHub 仓库地址、PR 编号/链接和补充说明，进行只读、证据驱动的 PR 审查：理解 PR 解决的问题、评估方案优劣与局限、默认优先使用内置“好 PR”标准并在必要时联网补充、识别恶意或高风险改动，并输出是否建议 merge 的 Markdown 决策报告。⚠️ 不适用：用户要你直接修改 PR 代码、直接 merge PR、或在本地执行 PR 分支中的不可信代码。

- 因本 skill 设计缺陷导致的 bug，先用 `bensz-collect-bugs` 规范记录到 `~/.bensz-skills/bugs/`，不要直接修改用户本地已安装的 skill 源码；若有 workaround，先记 bug，再继续完成任务。

- 因本 skill 设计缺陷导致的 bug，先用 `bensz-collect-bugs` 规范记录到 `~/.bensz-skills/bugs/`，不要直接修改用户本地已安装的 skill 源码；若有 workaround，先记 bug，再继续完成任务。

当需要把本仓库 pipelines/skills 下的所有 skills 安装到系统级（默认同时安装到 Codex: ~/.codex/skills 和 Claude Code: ~/.claude/skills），以便在任意项目/对话中可被发现与调用时使用。使用 MD5 哈希进行版本控制，仅安装有更新的 skills；支持强制覆盖安装、指定单一目标安装和远程安装模式（--remote --check/--auto）。

- 因本 skill 设计缺陷导致的 bug，先用 `bensz-collect-bugs` 规范记录到 `~/.bensz-skills/bugs/`，不要直接修改用户本地已安装的 skill 源码；若有 workaround，先记 bug，再继续完成任务。

基于 PTO-COMM ISA 开发通信算子的完整指南。涵盖 Host-Device 架构、文件结构、通信模式（P2P/集合通信/通算融合）、同步策略、信号矩阵设计、多 Block 调度、远端地址管理、构建系统配置等。触发：需要使用 PTO-COMM 开发通信算子、设计通信 kernel、编写 Host 侧代码、配置 CMakeLists 时。

PTO ISA 新增指令时需要更新的文档文件和位置模式总结。根据 docs/assembly/README.md 中的分类动态选择需要修改的汇编文件。触发：新增 PTO 指令（如 TPOW、TPOWS）后需要同步更新文档时。

Delegate CJK/Chinese content tasks to Gemini CLI. Use this skill for writing Chinese reports, Threads posts (美洲更新), comments, translations, and any task requiring native-quality Traditional Chinese output. Gemini handles the content generation while Claude plans, reviews, and integrates.

'Add support for a new target platform to acplugin (e.g., Windsurf, Zed, etc.)'

策略引擎、权限决策、RBAC、策略模式、责任链--- # Rust XACML - 策略引擎技能 > 本技能提供策略决策引擎的通用解决方案，包括 RBAC、责任链、策略模式等。 ## 核心概念 ### 1. 策略引擎架构 ``` 策略决策流程 ┌─────────────────────────────────────────────────────────────┐ │ 请求上下文 │ │ (主体、资源、操作、环境) │ └────────────────────────┬────────────────────────────────────┘ ▼ ┌─────────────────────────────────────────────────────────────┐ │ 策略决策点 (PDP) │ │ ├── 策略加载 │ │ ├── 策略评估 │ │ └── 决策组合 │ └────────────────────────┬────────────────────────────────────┘ ▼ ┌─────────────────────────────────────────────────────────────┐ │ 决策结果 │ │ (Permit / Deny / NotApplicable / Indeterminate) │ └─────────────────────────────────────────────────────────────┘ ``` ### 2. 决策类型 | 结果 | 含义 | 处理 | |-----|------|-----| | **Permit** | 允许访问 | 执行操作 | | **Deny** | 拒绝访问 | 返回 403 | | **NotApplicable** | 无匹配策略 | 使用默认规则 | | **Indeterminate** | 评估错误 | 返回 500 |

本文档面向后续 agent-code 与开发者，提供 VSeed 子包的架构、流程与关键模块的一页式上下文。

This file defines 25 specialized AI agents for Specification Driven Development across all platforms.

Makepad Skills Evolution workflow skill. Use this skill when the user needs This skill enables makepad-skills to self-improve continuously during development and the operator should preserve the upstream workflow, copied support files, and provenance before merging or handing off.

Makepad Skills Evolution workflow skill. Use this skill when the user needs This skill enables makepad-skills to self-improve continuously during development and the operator should preserve the upstream workflow, copied support files, and provenance before merging or handing off.

\u65c5\u884c\u5065\u5eb7\u5206\u6790\u6280\u80fd workflow skill. Use this skill when the user needs \u5206\u6790\u65c5\u884c\u5065\u5eb7\u6570\u636e\u3001\u8bc4\u4f30\u76ee\u7684\u5730\u5065\u5eb7\u98ce\u9669\u3001\u63d0\u4f9b\u75ab\u82d7\u63a5\u79cd\u5efa\u8bae\u3001\u751f\u6210\u591a\u8bed\u8a00\u7d27\u6025\u533b\u7597\u4fe1\u606f\u5361\u7247\u3002\u652f\u6301WHO/CDC\u6570\u636e\u96c6\u6210\u7684\u4e13\u4e1a\u7ea7\u65c5\u884c\u5065\u5eb7\u98ce\u9669\u8bc4\u4f30\u3002 and the operator should preserve the upstream workflow, copied support files, and provenance before merging or handing off.

ddi

Persona casting agent for rapid generation, persistence, lifecycle management, and inter-agent synchronization. Generates personas from diverse inputs, centrally manages them via a registry, evolves them data-driven, and distributes in a unified format to downstream agents.

Legal document review for Terms of Service, Privacy Policy, and Tokushoho compliance. Clause gap detection, risk flagging, and regulatory alignment. Use when reviewing or drafting legal documents. Don't use when legal advice is needed — consult a lawyer.

Competitive research, differentiation analysis, and strategic positioning. Feature matrices, SWOT analysis, benchmarking, positioning maps, battle cards, win/loss analysis, and LLM brand visibility. Research only — does not write code.

Engineer self-branding strategist that transforms technical contributions into a professional brand. Use when GitHub/LinkedIn/blog/conference/SNS positioning, profile optimization, or content strategy is needed.

Thinking refraction agent that challenges assumptions, combines cross-domain knowledge, and shifts perspectives to reframe problems. Use when breaking through stuck situations or paradigm shifts are needed. Does not write code.

Domain knowledge agent for Japanese income tax filing (kakutei shinkoku). Guides income classification, deduction optimization, tax calculation, and filing procedures for freelancers, sole proprietors, and side-business salary earners. Does not write code.

Mobile development specialist. Implements best practices for React Native/Flutter/SwiftUI/Jetpack Compose, navigation, offline support, and store review preparation. Use when mobile app development is needed.

Meta-orchestrator that coordinates specialist AI agent teams. Decomposes requests into minimum viable agent chains, spawns each agent as an independent session via Agent tool in AUTORUN modes, and drives to final output automatically. Use when multi-agent coordination is needed.

Pre-mortem analysis and failure mode enumeration agent. Systematically identifies failure scenarios for plans, designs, and features, scoring them with RPN/AP. Does not write code.

Multi-session parallel orchestrator using Claude Code Agent Teams API and Codex CLI Subagents to launch, manage, and coordinate concurrent task execution across multiple instances. Use when parallel work is needed.

User research specialist. Designs interview guides, usability test plans, qualitative data analysis, persona creation, and journey mapping. Complements Echo's UI validation. Use when user research design or analysis is needed.

Narrative design agent that tells product and feature use cases as customer-centric stories. Use when customer experience storytelling, scenario stories, or product narratives are needed.

Storybook story creation, catalog management, and Visual Regression integration. Handles UI component documentation, visual testing, and CSF 3.0/Factories format story creation. Supports Storybook 10 (ESM-only) and React Cosmos. Use when component documentation or visual testing is needed.

Scheduling and time-aware logic architect for cron, timezone/DST, retry/backoff, and business-calendar design. Use when schedule design is needed.

Session replay analysis, persona-based behavioral pattern extraction, and UX issue storytelling. A behavioral archaeologist who reads the 'why' from actual user operation logs. Collaborates with Researcher/Echo for persona validation.

User feedback collection, NPS survey design, review analysis, sentiment analysis, feedback classification, and insight extraction reports. Use when establishing feedback loops.

Tech blog/article series authoring for note/Zenn/Qiita/dev.to. Not for specs (Scribe) or microcopy (Prose).

Changelog 生成器 - 从 Git 历史自动生成 CHANGELOG

项目目录结构地图 - 生成带说明的可视化目录树，快速理解任意代码库

环境变量管理器：扫描、校验、同步 .env 文件，生成安全配置模板

编程错误消息翻译专家 — 将英文错误消息翻译成中文，解释原因并提供修复方案