core-test-failure-output
Run core PHP tests via tests/run_all.sh and return only failed test files with exact error/throw/fail lines. Use when user asks for failure-only raw output with no narrative summary.
来源信息
- 仓库
- LumiroCode/copilot-instructions
- 最近来源活动
- 2026年3月29日 14:35
- 检测到的 SKILL.md 语言
- 英语
- 星标
- 0
- Forks
- 0
安装方式
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
检查来源文件
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
文件资源管理器
2 个文件正在显示 SKILL.md
SKILL.md
来源说明 · 只读预览- name
- core-test-failure-output
- description
- Run core PHP tests via tests/run_all.sh and return only failed test files with exact error/throw/fail lines. Use when user asks for failure-only raw output with no narrative summary.
- argument-hint
- Optional scope or runner variant, e.g. core tests or package tests
- user-invocable
- true
- metadata
- {"skill":"core-test-failure-output","description":"Run core PHP tests and return only failed test files with exact failure lines, no summary.","usage":"Use when user requests raw failure output from core tests without narrative.","input":"Optional scope/runner variant, default is core suite via tests/run_all.sh.","output":"Exact lines from failure collector script, including NO_FAILED_TEST_FILES or FAILED_FILE blocks."}
# Core Test Failure Output
## When to use
- User asks to run core tests and show only failures.
- User requests exact failure lines, not a prose summary.
- User wants failed test file list plus raw failure messages.
## Inputs
- Default: run core suite from `./tests/run_all.sh`.
- Optional variants: adapt runner only if user explicitly requests another suite.
## Procedure
1. Run [failure collector script](./scripts/run_core_test_failure_output.sh).
2. Return script output as-is.
3. Do not add generated narrative, interpretation, or paraphrasing.
## Decision points
- If output is `NO_FAILED_TEST_FILES`, return that exact line only.
- If failures exist, return each `FAILED_FILE:` block with the captured lines.
- If the runner is missing/unexecutable, return the emitted error line unchanged.
## Completion checks
- Runner used is `./tests/run_all.sh`.
- Every reported failed file appears as `FAILED_FILE: <path>`.
- Failure body preserves exact lines produced by failed tests.
- No additional generated summary text is added.
在 GitHub 查看