AGENT SKILL

yq-toolchain-update

Audit or update this Windows computer's fixed third-party Codex helper toolchain when the user requests local helper-tool maintenance. Always show the complete registered list and obtain confirmation for each update group; exclude Codex itself and general runtimes unless explicitly added.

License 许可
MIT
所属仓库
EthanYoQ/Skill-hub
最近更新
2026-11-07
Stars
★ 11
Forks
5

01 前言 Front Matter

name
yq-toolchain-update
description
Audit or update this Windows computer's fixed third-party Codex helper toolchain when the user requests local helper-tool maintenance. Always show the complete registered list and obtain confirmation for each update group; exclude Codex itself and general runtimes unless explicitly added.

02 资源文件 2

03 SKILL.md 正文

YQ Toolchain Update

Maintain the fixed local helper-tool list in tool-registry.json. This Skill manages ordinary third-party command-line tools; use skill-lifecycle-manager separately for Skills' own sources, installs, and upgrades.

Required first response

Run scripts/toolchain-audit.ps1 -Mode Inventory before any update. Show the user the complete table of every selected tool, not only a count. Include its group, installed/not-installed state, observed version output, update source, and a short plain-language reason for the recommended order.

The registry is the single source of truth for the fixed list and its three groups: collaboration and cloud; browser, MCP, and search; content platforms. Show exact tool names only from the inventory output, so the entrypoint does not become a stale second copy.

Do not add Codex, Node, Python, Git, Docker, Bun, pnpm, uv, or other general runtimes unless the user explicitly asks.

Confirmation gates

  1. After showing the complete list and current state, get explicit approval for group 1.
  2. Before a group, show the full list for that group, available stable updates, and likely effect in non-technical language.
  3. Update one tool at a time with its source-native method from the registry. Verify its version and basic command before starting the next tool.
  4. At the end of a group, show a complete result table: updated, already current, skipped, or blocked. Stop and get explicit approval before the next group.
  5. Stop immediately for interactive login, authorization, configuration migration, or a failed validation. State exactly what needs the user's decision; never silently skip it or continue to another group.

Use stable releases only. Never run unbounded bulk updates such as a bare npm update -g or winget upgrade --all.

Safe update rules

  • Verify the official source and current stable version immediately before an update. Record the old and new version.
  • Preserve existing credentials, cookies, profiles, and user configuration. Do not print their contents.
  • Use package-manager readback plus a non-authenticated --version or --help check after each change.
  • If a PowerShell function or alias has the same name as a registry executable, report the conflict and use the verified application path for the probe. Do not run a shadowing function as if it were the registered tool.
  • For Playwright, update only the browser engines already installed unless the user asks for another browser. Do not alter the user's Chrome profile or login state.
  • Agent Reach can route to other tools but does not own their installations. Update its integrations as separate registry entries.
  • For LinkedIn, first search active Codex/Claude MCP configuration paths without printing secrets. Only replace a legacy server command after finding a real reference and receiving confirmation. If none is active, install and validate mcp-server-linkedin, leave the legacy command as a rollback path, and do not create a new MCP configuration.

Reporting

Keep reports simple. List every requested tool by name, including ones already current. Separate observed facts from unverified account health. End a completed three-group run by asking whether the user wants the registry or this Skill changed; do not change either automatically.

04 同仓库更多 Skills

EthanYoQ
EthanYoQ/Skill-hub

xhs-pharma-social-listening

Run evidence-disciplined Xiaohongshu social listening for foreign pharma employee pain points. Use when the user asks to collect and analyze notes/comments about pharma companies, aliases, roles, compliance, medical affairs, or AI topics, then produce a cited and scored sampled-insight report. Includes independent-evidence counting, MediaCrawler batch collection, and CAPTCHA/461 fallback to Agent Reach/OpenCLI.

Skill与智能体开发
★ 11
EthanYoQ
EthanYoQ/Skill-hub

healthcare-phi-compliance

Protected Health Information (PHI) and Personally Identifiable Information (PII) compliance patterns for healthcare applications. Covers data classification, access control, audit trails, encryption, and common leak vectors. Use when code touches PHI or PII in a healthcare system, or when auditing access control, audit trails, or leak vectors.

律师与法务Skill与智能体开发
★ 11
EthanYoQ
EthanYoQ/Skill-hub

healthcare-eval-harness

Patient safety evaluation harness for healthcare application deployments. Automated test suites for CDSS accuracy, PHI exposure, clinical workflow integrity, and integration compliance. Blocks deployments on safety failures. Use when a healthcare deployment must be gated on patient-safety tests for CDSS accuracy, PHI exposure, and workflow integrity.

Skill与智能体开发
★ 11
EthanYoQ
EthanYoQ/Skill-hub

healthcare-emr-patterns

EMR/EHR development patterns for healthcare applications. Clinical safety, encounter workflows, prescription generation, clinical decision support integration, and accessibility-first UI for medical data entry. Use when building EMR or EHR features such as encounter workflows, prescription generation, or clinical data entry UI.

Skill与智能体开发
★ 11
EthanYoQ
EthanYoQ/Skill-hub

healthcare-cdss-patterns

Clinical Decision Support System (CDSS) development patterns. Drug interaction checking, dose validation, clinical scoring (NEWS2, qSOFA), alert severity classification, and integration into EMR workflows. Use when building clinical decision support — drug interaction checks, dose validation, clinical scoring, or alert severity.

Skill与智能体开发
★ 11
EthanYoQ
EthanYoQ/Skill-hub

dxy-task-clarify

用户的请求含糊、缺背景、缺资料说明、缺读者或用途时使用。先用少量带选项的追问把任务问清楚,写成任务卡,用户确认后再执行。尤其适用于医生、科研人员和药企各岗位的工作,例如「帮我分析一下这些数据」「帮我准备明天的拜访」「写个总结」「做个 PPT」「回一下这封邮件」「整理一下会议纪要」「整理一下这些文献」「帮我准备下周的讲课」「写个病例汇报」「帮我写个提示词」。用户说「先问我」「帮我理清需求」「我不知道怎么描述」「我不会问 AI」时也使用。用途、资料、读者都已说清的请求,纯翻译,以及写代码、改代码,不使用。

Skill与智能体开发
★ 11