Lists multiple specific concrete actions: prepare Azure apps for deployment, infrastructure via Bicep/Terraform, azure.yaml, Dockerfiles, create/modernize apps. Also explicitly excludes cross-cloud migration, which adds clarity.

Clearly answers both 'what' (prepare Azure apps for deployment with infra Bicep/Terraform, azure.yaml, Dockerfiles) and 'when' (explicit WHEN clause with extensive trigger phrases). Also specifies when NOT to use it (cross-cloud migration), adding further completeness.

Excellent coverage of natural trigger terms users would say, including 'create app', 'build web app', 'deploy to Azure', 'generate Terraform', 'generate Bicep', 'function app', 'containerized Node.js app', 'static portfolio website', 'todo list with frontend and API', and many more variations.

Clearly scoped to Azure app creation/modernization and deployment preparation, explicitly distinguishing itself from the azure-cloud-migrate skill. The Azure-specific triggers and exclusion clause make it unlikely to conflict with other skills.

Extremely verbose and repetitive. The same instructions are stated multiple times — the plan-first requirement is repeated at least 4-5 times with nearly identical wording (Rules #1, the PLAN-FIRST WORKFLOW box, Phase 1 intro, Phase 1 Step 6). The mandatory hand-off to azure-validate is repeated 3+ times. The file is heavily padded with warnings, bold text, and ⛔/❌ markers that add visual noise without new information. Much of this content (e.g., explaining that skipping steps causes failures) could be stated once.

The skill provides a structured workflow with specific file paths, tool references, and clear step sequences, which is good. However, it contains no executable code examples, no concrete Bicep/Terraform snippets, and no sample deployment-plan.md content. Nearly all concrete guidance is deferred to reference files (analyze.md, requirements.md, etc.) which are not provided. The skill itself is more of a routing/orchestration document than an actionable guide.

The multi-phase workflow is clearly sequenced with explicit blocking points (STOP before Phase 2), mandatory validation checkpoints (validate before deploy, update plan status before hand-off), feedback loops (specialized routing checks), and a clear linear progression. Destructive action safeguards and error prevention are well-defined. The Phase 1 → approval gate → Phase 2 → validate → deploy chain is unambiguous.

The skill references many external files (references/analyze.md, references/requirements.md, references/architecture.md, etc.) which is good progressive disclosure structure. However, no bundle files were provided, so we cannot verify these references exist. The main file itself is quite long (~180+ lines) and contains significant repetition that could be consolidated. The SDK Quick References section and routing tables are well-organized, but the repeated mandatory warnings bloat the top-level document.