skill-creator: lifecycle guidance and Azure SDK patterns reference#379
Conversation
Split from #375: updates skill-creator SKILL.md with skill lifecycle guidance and adds references/azure-sdk-patterns.md. Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com> Copilot-Session: 9429ad33-6f31-4ec5-ba00-0f11b5afb82e
There was a problem hiding this comment.
Pull request overview
Expands Azure SDK skill authoring guidance with lifecycle, token-budget, validation, and reference-organization patterns.
Changes:
- Adds workflow and reference-extraction guidance.
- Introduces efficiency and regeneration validation gates.
- Documents the Azure SDK reference-directory structure.
Reviewed changes
Copilot reviewed 2 out of 2 changed files in this pull request and generated 10 comments.
| File | Description |
|---|---|
.github/skills/skill-creator/SKILL.md |
Adds lifecycle, authoring, validation, and regeneration guidance. |
.github/skills/skill-creator/references/azure-sdk-patterns.md |
Adds reference-directory and sync/async guidance. |
Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
Azure SDK guidance should not express a preference for either sync or async clients. Updated SKILL.md and azure-sdk-patterns.md to present both forms as first-class options with equal priority, removing the "pick one mode" directives while retaining the rule against mixing sync and async within a single call path.
The previous benchmark table referenced azure-ai-ml-py and other Python skills with descriptions that didn't match their actual content (e.g., azure-ai-ml-py covers workspaces/data/models/compute/jobs/pipelines, not a single train+register workflow). Replace with verified Rust skills from this repository that actually demonstrate the one-workflow/1-2-example criteria. Also drop the misleading 'Tokens Δ' column which implied historical optimization comparisons not applicable to new skills.
|
@copilot Fix the outstanding review comments. |
Fixed in commit I investigated the failing Smoke Test run and logs, and the failure was caused by an invalid Validated locally with the CI-equivalent commands ( |
| > 1. **Prefer `DefaultAzureCredential` for local development.** It works as-is with Azure CLI / VS Code / Developer CLI. For production, either constrain `DefaultAzureCredential` to production-safe credentials or use a specific credential directly. Avoid connection strings, account/API keys — they bypass Entra audit and rotation. | ||
| > - Local dev: `DefaultAzureCredential` works as-is. | ||
| > - Production: set `AZURE_TOKEN_CREDENTIALS=prod` (or `AZURE_TOKEN_CREDENTIALS=<specific_credential>`) to constrain the credential chain to production-safe credentials. |
| - Exceeding max limit → refactor into `/references/` subdirectories | ||
| - When approaching 500 lines → move entire sections to reference files | ||
| - Annotate with `<!-- Token Count: ~XXXX (target: 1100, max: 1500) -->` immediately below the skill's H1 |
| packages: | ||
| - . | ||
|
|
||
| onlyBuiltDependencies: | ||
| - esbuild |
Split from #375 (1 of 5).
Adds skill lifecycle guidance to skill-creator and a new Azure SDK patterns reference.
Files: