auth

Semua alur auth HARUS melalui AuthResolver. Tidak boleh menggunakan os.getenv() langsung untuk variabel token dalam kode aplikasi.

npx skills add https://github.com/microsoft/apm --skill auth

Auth Skill

Auth expert persona

When to activate

  • Any change to src/apm_cli/core/auth.py or src/apm_cli/core/token_manager.py
  • Code that reads GITHUB_APM_PAT, GITHUB_TOKEN, GH_TOKEN, ADO_APM_PAT
  • Code using git ls-remote, git clone, or GitHub/ADO API calls
  • Error messages mentioning tokens, authentication, or credentials
  • Changes to github_downloader.py auth paths
  • Per-host or per-org token resolution logic

Key rule

All auth flows MUST go through AuthResolver. No direct os.getenv() for token variables in application code.

Canonical reference

The full per-org -> global -> credential-fill -> fallback resolution flow is in docs/src/content/docs/getting-started/authentication.md (mermaid flowchart). Treat it as the single source of truth; if behavior diverges, fix the diagram in the same PR.

Bearer-token authentication for ADO

Azure DevOps Services (dev.azure.com, *.visualstudio.com) use this chain:

  1. ADO_APM_PAT env var if set
  2. AAD bearer via az account get-access-token --resource 499b84ac-1321-427f-aa17-267ca6975798 if az is installed and az account show succeeds
  3. Operation-time path-scoped git credential fill after HTTP 401/403 or a recognized Git authentication failure (is_ado_auth_failure_signal)
  4. Otherwise: AdoAuthChainExhaustedError from try_with_fallback, or build_error_context on resolve-time failures

Azure DevOps Server hosts (ADO_HOST / APM_ADO_HOSTS) skip az bearer: PAT, then fill.

ADO_APM_PAT is the env var name used by the auth flow. The AAD bearer source constant lives in src/apm_cli/core/token_manager.py as GitHubTokenManager.ADO_BEARER_SOURCE = "AAD_BEARER_AZ_CLI".

Stale-PAT fallback: on HTTP 401/403 or a recognized Git authentication failure, APM tries an available az bearer. A successful bearer retry emits:

[!] ADO_APM_PAT was rejected for {host}; fell back to az cli bearer.
[!]     Consider unsetting the stale variable.

Verbose source line (one per host, emitted under --verbose):

[i] dev.azure.com -- using bearer from az cli (source: AAD_BEARER_AZ_CLI)
[i] dev.azure.com -- token from ADO_APM_PAT

Diagnostic cases (build_error_context + AdoAuthChainExhaustedError in src/apm_cli/core/auth.py):

Initial PAT/bearer diagnostics do not describe the later helper retry.

  1. No PAT, no az: Azure DevOps requires authentication. You have two options -> install az and az login, or set ADO_APM_PAT.
  2. No PAT, az not signed in: same two-options copy with az login first.
  3. No PAT, wrong tenant: Your az cli session (tenant: ...) returned a bearer token, but Azure DevOps rejected it (HTTP 401). -> az login --tenant <correct-tenant>, or set ADO_APM_PAT.
  4. PAT set, request failed: ADO_APM_PAT is set, but the Azure DevOps request failed. -> rotate the PAT, az login on Services, or store a GCM credential.
  5. Helper empty or rejected: AdoAuthChainExhaustedError names only attempted steps -> refresh PAT, az login on Services, or store a repository credential in GCM. DNS, TLS, and timeout failures re-raise without unlocking fill or becoming auth exhaustion.
  6. Server, no PAT: Azure DevOps Server requires ADO_APM_PAT or a Git credential helper. -> set ADO_APM_PAT or store a GCM credential (az does not apply).

Lebih banyak skill dari microsoft

oss-growth
microsoft
Persona peretas pertumbuhan OSS
agent-framework-azure-ai-py
microsoft
Bangun agen Azure AI Foundry menggunakan Microsoft Agent Framework Python SDK (agent-framework-azure-ai). Gunakan saat membuat agen persisten dengan AzureAIAgentsProvider, menggunakan alat yang dihosting (code interpreter, file search, web search), mengintegrasikan server MCP, mengelola utas percakapan, atau mengimplementasikan respons streaming. Mencakup alat fungsi, keluaran terstruktur, dan agen multi-alat.
development
airunway-aks-setup
microsoft
Siapkan AI Runway di AKS — dari klaster kosong hingga model berjalan. Mencakup verifikasi klaster, instalasi controller, penilaian GPU, penyiapan penyedia, dan deployment pertama. KAPAN: "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".
devops
appinsights-instrumentation
microsoft
Panduan untuk instrumentasi aplikasi web dengan Azure Application Insights. Menyediakan pola telemetri, pengaturan SDK, dan referensi konfigurasi. KAPAN: cara menginstrumentasi aplikasi, SDK App Insights, pola telemetri, apa itu App Insights, panduan Application Insights, contoh instrumentasi, praktik terbaik APM.
devops
applicationinsights-web-ts
microsoft
Instrumentasi aplikasi browser/web dengan Application Insights JavaScript SDK (@microsoft/applicationinsights-web). Digunakan untuk Real User Monitoring (RUM) — tampilan halaman, klik, dependensi AJAX/fetch, pengecualian, peristiwa kustom, dan jejak agen GenAI sisi browser yang dikorelasikan dengan jejak OpenTelemetry backend. Mencakup pengaturan SDK Loader Script dan npm, ekstensi kerangka kerja (React, React Native, Angular), Click Analytics, inisialisasi telemetri, dan konvensi semantik OTel GenAI untuk span agen/alat/model yang dipancarkan dari browser.
devops
azure-ai-anomalydetector-java
microsoft
Bangun aplikasi deteksi anomali dengan Azure AI Anomaly Detector SDK untuk Java. Gunakan saat mengimplementasikan deteksi anomali univariat/multivariat, analisis deret waktu, atau pemantauan bertenaga AI.
development
azure-ai-language-conversations-py
microsoft
Implementasikan Pemahaman Bahasa Percakapan (CLU) menggunakan SDK Python azure-ai-language-conversations. Gunakan saat bekerja dengan ConversationAnalysisClient untuk menganalisis maksud dan entitas percakapan, membangun fitur NLP, atau mengintegrasikan pemahaman bahasa ke dalam aplikasi.
development
azure-ai-ml-py
microsoft
Azure Machine Learning SDK v2 untuk Python. Gunakan untuk ruang kerja ML, pekerjaan, model, kumpulan data, komputasi, dan pipeline. Pemicu: "azure-ai-ml", "MLClient", "ruang kerja", "registri model", "pekerjaan pelatihan", "kumpulan data".
development