documentation

作者: microsoft

hve-core 中審計、漂移、驗證與撰寫模式的標準文件能力。

npx skills add https://github.com/microsoft/hve-core --skill documentation

documentation

Overview

This skill provides the shared documentation capability used by the Documentation agent. It centralizes the durable knowledge for audit, drift, validate, and author flows in a single package so thin wrappers can load mode-specific guidance instead of embedding capability prose inline.

Mode map

The Documentation agent should load the relevant skill sections by mode:

ModePrimary load targetsNotes
auditreferences/conventions.md, references/coverage-method.md, references/validation-toolchain.md, references/content-sensitivity.mdReplaces the former Doc Ops workflow with a skill-driven audit loop.
driftreferences/conventions.md, references/code-doc-mapping.md, references/content-sensitivity.mdUses the repo-local mapping table and drift heuristics.
validatereferences/validation-toolchain.md, references/accessibility-checks.md, references/content-sensitivity.md, references/rai-guardrails.mdRuns docs validation and escalates formal review when needed.
authortemplates/guide.md, templates/reference.md, references/conventions.md, references/accessibility-checks.md, references/content-sensitivity.md, references/rai-guardrails.mdProduces narrative or reference docs with the repository's documented conventions.

Success criteria

For audit and validate modes, record results for all applicable local-safe checks and record existing CI status separately. Keep each CI-owned check that did not run as Pending CI, Skipped, Deferred, or Unavailable; do not report it as Passed.

Non-goals

This skill does not author ADRs, BRDs, PRDs, or other planning artifacts. It also does not embed accessibility, RAI, or security standards logic inline; it points to the dedicated reference files and routes formal assessment to the appropriate planner when the scenario requires it.

Working conventions

  • Prefer existing repository instructions and scripts over duplicated prose.
  • In audit and validate modes, run local-safe checks by default and summarize existing CI status separately. Invoke a CI-owned lane locally only when the task specifically asks to reproduce that named lane.
  • Treat browser installation, model or moderation environments, services, credentials, and interactive UI as separate prerequisites rather than automatic failure recovery.
  • Keep documentation changes factual and scoped to the current task.
  • Use the reference files below for mode-specific methods, heuristics, and checklists.
  • Escalate to planners for formal accessibility, RAI, or security review if the work requires a specialist assessment.

Stop rules

For audit and validate modes, stop after recording local-safe results and any separate CI evidence. Report missing specialized setup or required CI evidence separately; do not infer it, provision it automatically, or report it as passed.

Session tracking

Write session state to .copilot-tracking/documentation/ using a {{YYYY-MM-DD}}-session.md file for the run, following the standard session file conventions.

Reference files

  • references/conventions.md — Synthesis of the repository's markdown, writing-style, and Docusaurus conventions.
  • references/code-doc-mapping.md — Code-to-documentation mapping table and drift heuristics.
  • references/coverage-method.md — Gap and coverage analysis method for audit mode.
  • references/validation-toolchain.md — Validation command catalog and result interpretation.
  • references/accessibility-checks.md — Inline documentation checks and handoff triggers.
  • references/content-sensitivity.md — Pre-publish PII, secrets, confidentiality, and AI-disclosure checks.
  • references/rai-guardrails.md — Injection-boundary, attribution, human-review, and disclaimer guidance.

Templates

  • templates/guide.md — Structure template for narrative guides and how-to pages.
  • templates/reference.md — Structure template for reference and API-style documentation.

來自 microsoft 的更多技能

oss-growth
microsoft
開源增長駭客角色
agent-framework-azure-ai-py
microsoft
使用Microsoft Agent Framework Python SDK(agent-framework-azure-ai)构建Azure AI Foundry代理。适用于使用AzureAIAgentsProvider创建持久化代理、使用托管工具(代码解释器、文件搜索、网络搜索)、集成MCP服务器、管理对话线程或实现流式响应。涵盖函数工具、结构化输出和多工具代理。
development
airunway-aks-setup
microsoft
在AKS上設定AI Runway——從裸叢集到執行模型。涵蓋叢集驗證、控制器安裝、GPU評估、供應商設定及首次部署。時機:「設定AI Runway」、「上線AKS叢集」、「安裝AI Runway」、「airunway設定」、「部署模型至AKS」、「在AKS上進行GPU推論」、「在AKS上設定KAITO」、「在AKS上執行LLM」、「在AKS上使用vLLM」、「在AKS上設定模型服務」、「AI Runway控制器」。
devops
appinsights-instrumentation
microsoft
使用Azure Application Insights檢測Web應用程式的指南。提供遙測模式、SDK設定與組態參考。適用時機:如何檢測應用程式、App Insights SDK、遙測模式、什麼是App Insights、Application Insights指南、檢測範例、APM最佳實踐。
devops
applicationinsights-web-ts
microsoft
使用Application Insights JavaScript SDK(@microsoft/applicationinsights-web)為瀏覽器/Web應用程式進行檢測。適用於真實使用者監控(RUM)——頁面檢視、點擊、AJAX/fetch依賴、例外、自訂事件,以及與後端OpenTelemetry追蹤關聯的瀏覽器端GenAI代理追蹤。涵蓋SDK載入器指令碼與npm設定、框架擴充(React、React Native、Angular)、點擊分析、遙測初始化器,以及從瀏覽器發出的代理/工具/模型span的OTel GenAI語意慣例。
devops
azure-ai-anomalydetector-java
microsoft
使用適用於 Java 的 Azure AI 異常偵測器 SDK 建置異常偵測應用程式。在實作單變量/多變量異常偵測、時間序列分析或 AI 驅動監控時使用。
development
azure-ai-language-conversations-py
microsoft
使用 azure-ai-language-conversations Python SDK 實作對話語言理解(CLU)。當使用 ConversationAnalysisClient 分析對話意圖與實體、建置 NLP 功能,或將語言理解整合至應用程式時使用。
development
azure-ai-ml-py
microsoft
Azure Machine Learning SDK v2 for Python。用於機器學習工作區、作業、模型、資料集、計算資源與管線。 觸發詞:「azure-ai-ml」、「MLClient」、「workspace」、「model registry」、「training jobs」、「datasets」。
development