Builds, tests, and deploys Microsoft 365 apps and agents for Teams and Copilot. Includes sub-skills for project creation, local testing, cloud deployment, troubleshooting, and Slack-to-Teams migration. USE FOR: Teams agent, bot, tab, message extension, Declarative Agents, Custom Engine Agents, local testing, Agents Playground, Azure resource provision, remote deployment, Slack to Teams migration, cross-platform bot development, Block Kit to Adaptive Cards conversion. DO NOT USE FOR: general web development, non-bot/non-Teams projects.
中文简介正在准备中。当前显示原文说明。
查看安装方法安装前可以查看交给 AI 代理的指令内容。
Build Microsoft 365 agents and Teams apps using the ATK CLI.
Testing Strategy: Recommend Agents Playground first (faster, no M365 needed). Use Teams workflow only if user explicitly requests it.
Environment Variables: NEVER hardcode secrets or make up placeholder values. Always ask users for real values.
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-provisionTENANT_ID in .localConfigs → SDK uses wrong token authority → 401 from Bot ConnectorLong-Running Commands — WAIT for completion:
atk new, atk provision, atk deploy can take several minutesLocal Service Startup — Hangs terminal (expected):
npm run dev, npm start, python app.py, devtunnel host, etc. will hang — the process keeps running indefinitelyisBackground=true) — NEVER use isBackground=false for these commandsMonitor App Logs: Periodically check background terminal output for runtime errors. If the app crashes, read the error, fix the root cause, and restart.
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 --version # Must be > 1.1.5-beta
If ATK is not found or version is too old:
npm i -g @microsoft/m365agentstoolkit-cli@beta
| Option | Meaning | Recommendation |
|---|---|---|
-i | Interactive mode | Always use -i false in automation to avoid hanging |
-f | Project folder | Default 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. |
-h | Command help | Use atk <command> -h for quick syntax checks |
| Sub-Skill | When to Use | Reference |
|---|---|---|
| create-project | Scaffold new project from template, choose template, atk new | create-project/create-project.md |
| test-playground | Test locally with Agents Playground, agentsplayground, quick testing | test-playground/test-playground.md |
| test-teams | Run on Teams, devtunnel, sideload, Teams testing, test in Copilot | test-teams/test-teams.md |
| provision-deploy | Provision Azure resources, deploy to cloud, atk provision, atk deploy | provision-deploy/provision-deploy.md |
| troubleshoot | Fix errors, 401, port conflicts, YAML errors, stale bots | troubleshoot/troubleshoot.md |
| slack-to-teams | Migrate Slack bot to Teams, cross-platform bridging, Block Kit to Adaptive Cards | slack-to-teams/SKILL.md |
MANDATORY: Before executing any workflow, read the corresponding sub-skill document.
Match user intent to the smallest valid workflow.
| User Intent | Workflow (read in order) |
|---|---|
| Build new app from scratch | create-project → test-playground |
| Test existing project locally | test-playground (recommended) or test-teams |
| Deploy to Azure | provision-deploy |
| Fix broken bot | troubleshoot → re-test |
| Migrate Slack bot to Teams | slack-to-teams |
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.
Resolve config values only when missing. If a value is already known in the session, reuse it.
If m365agentstoolkit*.yml exists in the current folder, treat it as an ATK project and parse configuration.
Resolve variables referenced in m365agentstoolkit*.yml. Common variables:
AZURE_OPENAI_API_KEY
AZURE_OPENAI_ENDPOINT
AZURE_OPENAI_DEPLOYMENT_NAME
If required values are missing, ask the user for only the missing ones.
Refer to manifest-and-yaml.md for full config-file details.
还没有评价。欢迎分享你的使用感受。
无需登录即可发表。欢迎分享使用感受、提问，或说明在哪些环境下无法运行。
还没有评论。来写第一条吧。
Build Azure AI Foundry agents using the Microsoft Agent Framework Python SDK (agent-framework-azure-ai). Use when creating persistent agents with AzureAIAgentsProvider, using hosted tools (code interpreter, file search, web search), integrating MCP servers, managing conversation threads, or implementing streaming responses. Covers function tools, structured outputs, and multi-tool agents.
中文简介正在准备中。当前显示原文说明。
Set up AI Runway on AKS — from bare cluster to running model. Covers cluster verification, controller install, GPU assessment, provider setup, and first deployment. WHEN: "setup AI Runway", "onboard AKS cluster", "install AI Runway", "airunway setup", "deploy model to AKS", "GPU inference on AKS", "KAITO setup on AKS", "run LLM on AKS", "vLLM on AKS", "set up model serving on AKS", "AI Runway controller".
中文简介正在准备中。当前显示原文说明。
Diagnose Day-2 AKS GPU and KAITO incidents using profile-aware, read-only evidence. WHEN: 'Insufficient nvidia.com/gpu', GPU pod Pending, model-load OOM, DCGM/VRAM, KAITO Workspace not ready, or GPU autoscaling. DO NOT USE FOR: setup (airunway-aks-setup), non-GPU incidents (aks-troubleshooting), standalone VM quota (azure-quotas), or generic cost (cost-analysis or cost-optimization from the optional azure-cost plugin).
中文简介正在准备中。当前显示原文说明。
Lookup documented AKS fixes only when the prompt includes an exact catalog signature and all of its qualifiers: VMCannotFitEphemeralOSDisk; NodePoolMcVersionIncompatible; 'NodeImageVersion is not accepted'; AKS SkuNotAvailable with size, location, and zone; ZonalAllocationFailed with insufficient zone capacity; OverconstrainedAllocationRequest with listed constraints; nested AKS vmssCSE/CSE VMExtensionError_OutboundConnFail, VMExtensionError_K8SAPIServerConnFail, or VMExtensionError_K8SAPIServerDNSLookupFail; or AllocationFailed with the full cataloged internal-error or insufficient-regional-capacity message. Never use for quota errors, code-only or bare wrappers, generic symptoms, incomplete signatures, or failures outside AKS; use aks-troubleshooting or azure-diagnostics.
中文简介正在准备中。当前显示原文说明。
Collects bounded packet captures from AKS nodes and Azure network configuration for wire-level evidence. WHEN: "capture packets on an AKS node", "take a pcap", "run tcpdump on AKS", "prove where packets drop". Use for explicit packet-capture intent after read-only diagnostics, not general AKS connectivity or ingress troubleshooting.
中文简介正在准备中。当前显示原文说明。
Debug live Azure Kubernetes Service (AKS) incidents with a read-only, evidence-first investigation. WHEN: pod crashes or Pending, CrashLoopBackOff, OOMKilled, ImagePullBackOff, node NotReady, DNS or ingress failure, connectivity timeout, network policy, SNAT exhaustion, node-pool scaling blocked by QuotaExceeded or InsufficientVCPUQuota, upgrade stuck, spot or zone disruption, a bare VMExtensionProvisioningError or AllocationFailed wrapper, an uncataloged capacity symptom, or 'investigate my AKS cluster'. DO NOT USE FOR: packet capture (use aks-network-capture); GPU or model-serving issues (use aks-gpu-inference); cluster creation or provisioning (use azure-kubernetes); cost (use cost-analysis from the optional azure-cost plugin); pod rightsizing (use azure-kubernetes); a fully qualified documented AKS signature with every required nested qualifier (use aks-known-issues); standalone failures on non-AKS Azure resources (use azure-diagnostics). Unqualified errors and open-ended incidents stay here only for AKS.
中文简介正在准备中。当前显示原文说明。