teams-app-developer

作者: microsoft

使用 M365 Agents Toolkit CLI 建置、測試和部署基於程式碼的 Teams 應用程式。適用於:自訂引擎代理 (CEA)、Teams 機器人、索引標籤、訊息擴充功能…

npx skills add https://github.com/microsoft/work-iq --skill teams-app-developer

name: teams-app-developer description: > Build, test, and deploy code-based Teams apps using the M365 Agents Toolkit CLI. USE FOR: Custom Engine Agents (CEA), Teams bots, tabs, message extensions, Agents Playground local testing, Azure provision/deploy, Slack-to-Teams migration, cross-platform bot development, Block Kit to Adaptive Cards conversion, AI model integration (OpenAI/Azure/Anthropic/Bedrock). DO NOT USE FOR: declarative agents — use the declarative-agent-developer skill instead. Triggers: "build a teams bot", "custom engine agent", "CEA", "teams agent", "tab app", "message extension", "test with agents playground", "provision to azure", "deploy to azure", "migrate slack bot", "slack to teams", "convert block kit", "add azure openai to my bot"

Teams App Developer Skill

Build code-based Microsoft 365 agents and Teams apps (CEA, bot, tab, message extension) using the ATK CLI.

Declarative Agents (DA): Use the declarative-agent-developer skill — it owns all DA workflows including scaffolding, manifest editing, capability config, API/MCP plugins, OAuth, localization, and deployment. This skill covers code-based project types only.

AI Behavior Guidelines

  1. Testing Strategy: Recommend Agents Playground first (faster, no M365 needed). Use Teams workflow only if user explicitly requests it.

  2. Environment Variables: NEVER hardcode secrets or make up placeholder values. Always ask users for real values.

  3. Error Handling: Read error messages carefully. Check env/.env.local, .localConfigs, and atk auth list. Common pitfalls:

    • AADSTS7000229 → aadApp/create missing generateServicePrincipal: true in YAML — add it and re-provision
    • Missing TENANT_ID in .localConfigs → SDK uses wrong token authority → 401 from Bot Connector
    • 401 persists after auth fix → devtunnel URL may be blacklisted — create a fresh tunnel
    • See troubleshoot/troubleshoot.md for full diagnostic steps
  4. Long-Running Commands — WAIT for completion:

    • atk new, atk provision, atk deploy can take several minutes
    • Always wait for completion before running the next step (timeout 120000ms+)
  5. Local Service Startup — Hangs terminal (expected):

    • npm run dev, npm start, python app.py, devtunnel host, etc. will hang — the process keeps running indefinitely
    • ALWAYS run as a background process (isBackground=true) — NEVER use isBackground=false for these commands
    • Do NOT wait for it to "finish" — verify startup by checking output for "listening on port" or tunnel URL
    • If errors appear, read logs, diagnose, fix, restart
    • Use a NEW terminal to launch Agents Playground or open Teams sideloading URL
  6. Monitor App Logs: Periodically check background terminal output for runtime errors. If the app crashes, read the error, fix the root cause, and restart.

  7. Telemetry Tagging: Before running any atk CLI commands, set the session environment variable so all CLI invocations are tagged as skill-initiated:

    export ATK_CLI_SKILL=true
    

    Run this once at the start of the session. All subsequent atk commands in the same terminal will inherit it.

ATK CLI Setup

Before any ATK commands, verify the CLI is available:

atk --version  # Must be > 1.1.5-beta

If ATK is not found or the version is too old → use the install-atk skill to install or update it, then return here to continue.

CLI Global Options

OptionMeaningRecommendation
-iInteractive modeAlways use -i false in automation to avoid hanging
-fProject folderDefault to be current directory, used when specifying a custom folder. When scaffolding a new project, this is the parent folder where the project folder will be created under.
-hCommand helpUse atk <command> -h for quick syntax checks

Sub-Skills

Sub-SkillWhen to UseReference
create-projectScaffold new project from template, choose template, atk newcreate-project/create-project.md
test-playgroundTest locally with Agents Playground, agentsplayground, quick testingtest-playground/test-playground.md
test-teamsRun on Teams, devtunnel, sideload, Teams testing, test in Copilottest-teams/test-teams.md
provision-deployProvision Azure resources, deploy to cloud, atk provision, atk deployprovision-deploy/provision-deploy.md
troubleshootFix errors, 401, port conflicts, YAML errors, stale botstroubleshoot/troubleshoot.md
slack-to-teamsMigrate Slack bot to Teams, cross-platform bridging, Block Kit to Adaptive Cardsslack-to-teams/SKILL.md

MANDATORY: Before executing any workflow, read the corresponding sub-skill document.

Workflow Chains

Match user intent to the smallest valid workflow.

User IntentWorkflow (read in order)
Build new CEA/bot/tab/ME from scratchcreate-project → test-playground
Test existing project locallytest-playground (recommended) or test-teams
Deploy to Azureprovision-deploy
Fix broken bottroubleshoot → re-test
Migrate Slack bot to Teamsslack-to-teams
Create Declarative AgentUse declarative-agent-developer skill

MANDATORY: Before executing any slack-to-teams workflow, read slack-to-teams/SKILL.md first. The sub-skill contains a routed expert system with 100+ micro-expert files for cross-platform bot development.

Shared References

  • manifest-and-yaml.md — Project files, YAML config, env vars, .localConfigs flow
  • commands.md — ATK CLI commands: package, validate, share, collaborate
  • templates.md — Complete template catalog with language support
  • experts/ — 100+ micro-expert files: Teams SDK, Slack SDK, cross-platform bridging, deploy, AI models, security, language conversion
  • docs/ — Platform comparison guides: UI, messaging, identity, infrastructure, feature gaps

ATK Project Context Resolution

Resolve config values only when missing. If a value is already known in the session, reuse it.

Step 1: Detect ATK Project

If m365agentstoolkit*.yml exists in the current folder, treat it as an ATK project and parse configuration.

Step 2: Resolve Common Configuration

Resolve variables referenced in m365agentstoolkit*.yml. Common variables: AZURE_OPENAI_API_KEY AZURE_OPENAI_ENDPOINT AZURE_OPENAI_DEPLOYMENT_NAME

Step 3: Collect Missing Values

If required values are missing, ask the user for only the missing ones.

Refer to manifest-and-yaml.md for full config-file details.

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