菜单
Triggers the QA Test Runner subagent to reproduce failures and interpret logs.
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
Skill for reviewing Magic Modules schemas. Currently covers the Enum vs. String decision via the knowledge base.
Opt resource into MMv1 list-resource generation by setting `generate_list_resource: true`, validate it locally, and open a one-resource PR against GoogleCloudPlatform/magic-modules. Invoke when the user asks to add list-resource support for a specific MMv1 resource or to enable `generate_list_resource` for an eligible resource.
Fallback workflow for general implementation and debugging tasks that do not involve creating a new resource.
Workflow specifically for creating a new resource, supporting both autogen and manual generation.
Create a Pull Request (PR) against GoogleCloudPlatform/magic-modules following repository standards, including branch management, commit formatting, mandatory release notes, pre-PR verification checks, and gh CLI commands.
Validate that changes to the generated providers don't introduce breaking changes or fields with missing tests or missing documentation. Use this skill when the user wants to check for breaking changes, missing tests, or missing documentation, or other problems in the downstream providers.
| name | qa-test-runner |
| description | Triggers the QA Test Runner subagent to reproduce failures and interpret logs. |
qa-test-runnerUse this skill to delegate heavy test reproduction and log parsing to a dedicated Subagent. This keeps your main context window clean.
Gather context on what changed (the resource, the field, the PR) and which test you need to run.
Use the invoke_subagent tool. Pass the context of the change as part of the trigger prompt so the subagent knows what it's looking for!
Trigger Example: invoke_subagent("qa-test-runner", "Run <TEST_NAME>. We just made the following change: <CHANGE_DESCRIPTION>. Verify if it fails or passes. Use your active workspace to find the magic-modules path <PATH_TO_MAGIC_MODULES> and output debug logs to <PATH_TO_MAGIC_MODULES>/debug_output.")
Read the human-readable Markdown report returned by the subagent. Formulate a fix strategy using .agents/skills/operations/troubleshooting_reference.md based on their analysis.