api-changes

bởi microsoft

Sử dụng khi có thay đổi API ảnh hưởng đến khách hàng — tức là các tệp báo cáo API .md khác với nhánh chính. Hướng dẫn gán thẻ phát hành, đánh giá của Hội đồng API…

npx skills add https://github.com/microsoft/fluidframework --skill api-changes
Before doing any work, create one task/todo item per applicable step using your available task tooling (TaskCreate for Claude, TodoWrite for Copilot). Mark each task in_progress when you start it and completed when you finish. This prevents steps from being silently skipped as context grows.

API Changes Review

Step 1: Identify what changed

Read .claude/skills/comparison-base/SKILL.md and execute it with HEAD as $REVIEW_REF. This is required; use the target and comparison commit it resolves.

Compare the selected base with the working tree, not just HEAD, because this skill is often called immediately after API reports have been regenerated and those edits may be uncommitted:

git diff "$BASE_COMMIT" -- ':(glob)**/api-report/*.md'
git ls-files --others --exclude-standard -- ':(glob)**/api-report/*.md'

Treat the union of both command outputs as the changed API reports. For each untracked report, read the full file as a new API surface because it has no Git diff until staged.

Build a summary table and present it to the user:

PackageChange typeTag(s)Breaking?

Change types: addition, removal, signature change, tag promotion.

If all changes are @internal-only, tell the user there are no customer-facing API changes and stop.


Step 2: Check release tags, documentation, and export reachability

For any new exports, verify each has a release tag and flag any missing ones to the user — API Extractor will fail with ae-missing-release-tag. Help the user choose the right tag:

TagWhen to use
@publicStable, production-ready. Full SemVer. Use only when the shape is final.
@betaSeeking feedback, path to @public. Production OK with caution.
@alphaExperimental, early feedback only. Not for production. No stability guarantees.
@internalFramework-internal only, not for external consumers.

When in doubt: @alpha — easier to promote than demote. @legacy is a paired modifier (@legacy @public or @legacy @alpha) for FF v1 APIs; don't apply it to new APIs.

For every new customer-facing export (@public, @beta, @alpha) that is intended to be usable by package consumers, verify it is reachable from the package's public entrypoint, not just exported from the adjacent module or folder. Trace and update the export chain through every relevant index.ts barrel up to the package root entrypoint (typically src/index.ts, or tiered entrypoints such as src/alpha.ts / src/beta.ts where used). Missing parent-barrel exports are incomplete API changes. API Extractor may not report the intended API at all, and consumers are expected to import from the package's top-level entrypoint rather than reaching into subpaths.

Also check that each new customer-facing export (@public, @beta, @alpha) has TSDoc documentation — at minimum a summary, @param tags, and @returns if applicable. Flag any missing documentation to the user.


Step 3: Inform the user about API Council review

Tell the user whether their change requires API Council approval:

Changed surfaceApproval required?
@public, @legacy @public, @beta, @legacy @alphaYes — fluid-cr-api will be automatically assigned as a required reviewer on the PR
@alpha only (not @legacy)No — but early engagement with the council is encouraged
@internal onlyNo

Tell the user: council approval is a separate sign-off from the area owner review. To engage the council, they can reach out to the API Council member on their EM team or tag @FF API on Teams. Share this link with the user for more details: https://eng.ms/docs/experiences-devices/opg/office-shared/fluid-framework/fluid-framework-internal/fluid-framework/docs/dev/resources/api-council


Step 4: Assess breaking changes

A breaking change removes or modifies an existing API in a way that causes compile errors for consumers upgrading.

@public / @legacy+@public

If this is a breaking change to @public or @legacy @public, tell the user this is likely a mistake — major releases happen very rarely. Breaking @public APIs must be coordinated with a major release; the old API must be deprecated at least 3 months prior in a minor release with a clear replacement.

Share these links with the user for the required process:

@beta / @legacy+@alpha

If this is a breaking change to @beta or @legacy @alpha, tell the user:

  • Breaking changes may only land in minor versions that are an increment of 10 (2.10, 2.20, 2.30, …)
  • The PR must be staged on a test/breaks/client/#.#0/ branch and held until the break window opens
  • They should check whether partners (e.g. office-bohemia) consume the API directly — if in doubt, assume they do and allow 12 weeks lead time

Share these links with the user:

@alpha only

Tell the user: while @alpha has no contractual stability guarantees, there is an informal agreement not to break office-bohemia. If this change could break office-bohemia, it should be staged using the same process as above.

Also tell the user: if they skip this check and the change does break office-bohemia, the daily integration pipeline will catch it and FF OCE will revert the PR or contact them to do so ASAP.


Step 5: Deprecation checklist

If any API is being deprecated, check that the following are in place and flag anything missing to the user:

  • @deprecated TSDoc comment includes: version deprecated, version of removal, replacement, and a link to the tracking issue:
    /**
     * @deprecated 2.x.y. Removed in 3.0.0. Use {@link replacementApi} instead.
     * See {@link https://github.com/microsoft/FluidFramework/issues/ABCD} for context.
     */
    
  • GitHub issue filed using the "Deprecated API" template as a sub-issue of the appropriate tracking issue
  • In-codebase uses removed (test-only uses may remain with an explanatory comment)
  • Release documentation present (see Step 6)

Share this link with the user for full deprecation guidance: ../../../docs/content/Contributing/API-Deprecation.md


Step 6: Release documentation

Document customer-facing API changes: additions, modifications, deprecations, release level promotions, and removals. Follow the release-group guidance to choose the documentation format.

For groups that use changesets, check the branch diff and working tree for an existing entry in the group's .changeset directory. If a required changeset is missing, create one from the repo root:

pnpm flub changeset add --releaseGroup <releaseGroup> --empty

This creates a randomly named file in the selected release group's .changeset directory. Edit it with content based on what changed. In the YAML front matter, list only packages from that group where the change matters to consumers. Use bump type minor and a __section value from the release note sections.

Summary line rules (from .changeset/README.md): succinct, no terminal punctuation, no backtick formatting, present tense. Prefix test: mentally prepend "In this release," to verify it reads naturally. Body may include a code example for features, deprecations, and breaking changes.

After drafting the changeset, show the content to the user and confirm it looks right before moving on.


Step 7: Summary

Present the user with a clear summary:

  1. API changes found (table from Step 1)
  2. Any missing release tags or documentation
  3. Whether API Council review is required
  4. Any breaking change warnings and the process the user needs to follow
  5. Any deprecation issues
  6. Release documentation status

End with a clear go/no-go: "Your changes look good to merge" or "Please resolve these issues before merging: …"

Thêm skills từ microsoft

oss-growth
microsoft
Cá tính tăng trưởng OSS
agent-framework-azure-ai-py
microsoft
Xây dựng các tác nhân Azure AI Foundry bằng SDK Python của Microsoft Agent Framework (agent-framework-azure-ai). Sử dụng khi tạo các tác nhân bền vững với AzureAIAgentsProvider, sử dụng các công cụ được lưu trữ (trình thông dịch mã, tìm kiếm tệp, tìm kiếm web), tích hợp máy chủ MCP, quản lý chuỗi hội thoại hoặc triển khai phản hồi phát trực tuyến. Bao gồm các công cụ hàm, đầu ra có cấu trúc và các tác nhân đa công cụ.
development
airunway-aks-setup
microsoft
Thiết lập AI Runway trên AKS — từ cụm trống đến mô hình đang chạy. Bao gồm xác minh cụm, cài đặt controller, đánh giá GPU, thiết lập nhà cung cấp và triển khai đầu tiên. KHI NÀO: "thiết lập AI Runway", "onboard cụm AKS", "cài đặt AI Runway", "thiết lập airunway", "triển khai mô hình lên AKS", "suy luận GPU trên AKS", "thiết lập KAITO trên AKS", "chạy LLM trên AKS", "vLLM trên AKS", "thiết lập phục vụ mô hình trên AKS", "AI Runway controller".
devops
appinsights-instrumentation
microsoft
Hướng dẫn để instrument các ứng dụng web với Azure Application Insights. Cung cấp các mẫu telemetry, thiết lập SDK, và tài liệu tham khảo cấu hình. KHI NÀO: cách instrument ứng dụng, App Insights SDK, các mẫu telemetry, App Insights là gì, hướng dẫn Application Insights, ví dụ instrumentation, các phương pháp tốt nhất APM.
devops
applicationinsights-web-ts
microsoft
Instrument các ứng dụng trình duyệt/web bằng SDK JavaScript Application Insights (@microsoft/applicationinsights-web). Dùng cho Real User Monitoring (RUM) — lượt xem trang, nhấp chuột, phụ thuộc AJAX/fetch, ngoại lệ, sự kiện tùy chỉnh và dấu vết tác nhân GenAI phía trình duyệt tương quan với dấu vết OpenTelemetry phía backend. Bao gồm thiết lập SDK Loader Script và npm, tiện ích mở rộng framework (React, React Native, Angular), Click Analytics, trình khởi tạo telemetry và quy ước ngữ nghĩa OTel GenAI cho các span tác nhân/công cụ/mô hình phát ra từ trình duyệt.
devops
azure-ai-anomalydetector-java
microsoft
Xây dựng ứng dụng phát hiện bất thường với Azure AI Anomaly Detector SDK cho Java. Sử dụng khi triển khai phát hiện bất thường đơn biến/đa biến, phân tích chuỗi thời gian hoặc giám sát hỗ trợ AI.
development
azure-ai-language-conversations-py
microsoft
Triển khai Conversational Language Understanding (CLU) bằng SDK Python azure-ai-language-conversations. Sử dụng khi làm việc với ConversationAnalysisClient để phân tích ý định và thực thể trong hội thoại, xây dựng tính năng NLP, hoặc tích hợp hiểu ngôn ngữ vào ứng dụng.
development
azure-ai-ml-py
microsoft
Azure Machine Learning SDK v2 cho Python. Dùng cho không gian làm việc ML, công việc, mô hình, tập dữ liệu, tính toán và quy trình. Kích hoạt: "azure-ai-ml", "MLClient", "không gian làm việc", "đăng ký mô hình", "công việc đào tạo", "tập dữ liệu".
development