菜单
Use when defining where human judgment is still required, how exploratory work is captured, and which repeated checks should graduate into automation. Provides templates for decision rubrics, exploratory charters, regression checklists, and defect evidence.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Use when designing Copilot agents, authoring agent definitions, creating skill folders, or scaffolding instruction files. Provides templates and conventions for the agent ecosystem.
Use when designing, reviewing, or evolving API contracts. Provides OpenAPI templates, breaking-change checklists, versioning decision trees, and governance checklists.
API authentication, authorization, input validation, rate limiting, and protection patterns
Scan legacy applications to discover project structure, NuGet/npm dependencies, database connection strings, external service bindings, framework versions, and migration complexity scores. Produces structured JSON, YAML, or Markdown inventory reports.
Use when designing systems, recording architecture decisions, evaluating technologies, or assessing architectural risks. Provides C4 diagram templates, ADR format, technology selection matrices, and risk registers.
Deploy, scale, and manage containerized applications on Azure Container Apps with Dapr, revision management, and advanced networking
| name | manual-test-strategy |
| description | Use when defining where human judgment is still required, how exploratory work is captured, and which repeated checks should graduate into automation. Provides templates for decision rubrics, exploratory charters, regression checklists, and defect evidence. |
| compatibility | ["VS Code","Cursor","Windsurf","Claude Code"] |
| metadata | {"category":"Uncategorized","tags":["uncategorized"],"maturity":"beta","audience":["developers"]} |
| allowed-tools | ["bash","git","grep","find"] |
Use this skill when a feature, risk inventory, or change needs a structured decision about what stays manual, what gets automated next, and how evidence moves between both.
Agents should load the relevant template from this skill's directory before producing artifacts:
| Template | When to use |
|---|---|
rubric-template.md | Classifying behaviors as manual-only, automate-now, or hybrid |
charter-template.md | Structuring a time-boxed exploratory session |
checklist-template.md | Building a repeatable manual regression checklist |
defect-template.md | Capturing defect evidence for filing, triage, or automation handoff |
rubric-template.md.defect-template.md for any found defects.gh issue create pattern defined in the relevant agent.