界面已本地化;技能介绍与说明文档按所选语言由 AI 自动翻译,代码、命令、路径和专业术语保留原文,以原作者文档为准。
探索技能
数据与研究 · SKILL.md

validation-testing

validation-testing

【数据与研究】Selects and runs Scientific Agent Skills repository validation and tests: skills-ref validate, repo-specific metadata checks, tests/_meta, per-skill pytest suites, one-skill-per-process rules, tests/run_all.py, --isolated environments, and tests/skill-requirements.toml entries.

独立 GitHub 快照 · 自动分类 · 以原作者文档为准

技能说明

GitHub

name: validation-testing description: "Selects and runs Scientific Agent Skills repository validation and tests: skills-ref validate, repo-specific metadata checks, tests/_meta, per-skill pytest suites, one-skill-per-process rules, tests/run_all.py, --isolated environments, and tests/skill-requirements.toml entries." disable-model-invocation: true metadata: disco-role: operating license: MIT

Validation and Testing

Use this route when a task asks how to validate, test, or diagnose CI for skill changes in this repository.

First checks

  1. Install repository tooling with Python 3.13:

    uv sync --python 3.13
    
  2. Validate the changed skill's Agent Skills spec:

    uv run skills-ref validate skills/<name>
    
  3. Run the repo-wide structural guard after local issues are fixed:

    uv run --with pytest python -m pytest tests/_meta -q
    

Read references/validation-and-ci.md for CI-equivalent checks and references/test-runner.md for per-skill and isolated suite behavior. Use scripts/plan_skill_checks.py to inspect one skill and print an ordered command plan without executing tests.

Test-suite rules

  • Tests never live under skills/<name>/.

  • A skill that ships files under scripts/ needs tests/<name>/ with at least one test_*.py and a [skills.<name>] entry in tests/skill-requirements.toml.

  • Run one skill's suite in one Python process:

    uv run --with pytest python -m pytest tests/<name> -q
    
  • For isolated dependency environments, use:

    python tests/run_all.py --isolated <name>
    
  • Do not collect multiple skill suites with one pytest process; script module names intentionally collide across skills.

Escalation guidance

Change typeMinimum validation
SKILL.md prose onlyskills-ref validate skills/<name> and consider tests/_meta for links/line count.
Skill references/assets changedskills-ref validate, tests/_meta, and any touched examples/manual checks.
Skill scripts changed or addedskills-ref validate, tests/<name>, tests/run_all.py --isolated <name>, and tests/_meta.
Shared test contract or runner changedtests/_meta and python tests/run_all.py; use isolated selection for affected script-bearing skills.
CI workflow changedMirror the workflow commands locally and inspect path filters/permissions/timeouts.

Troubleshooting

Read references/troubleshooting.md when tests/_meta reports missing suites, skill-requirements.toml entries, broken local links, local paths, or unsafe script patterns.

我的收藏