AgentSkillsCN

run-all-ui-tests

依次运行 URLS4IRL 的所有 UI 测试。当用户被要求运行所有 UI 测试、运行完整的 UI 测试套件，或验证所有功能测试/Selenium 测试均通过时，可使用此技能。测试失败结果会被记录至 tmp/ 目录，并以带时间戳的文件形式保存；若测试全部成功，则自动清理相关文件。

SKILL.md
--- frontmatter
name: run-all-ui-tests
description: Run ALL UI tests for URLS4IRL sequentially. Use when asked to run all UI tests, run the full UI test suite, or verify all functional/Selenium tests pass. Records failures to tmp/ with timestamped files; cleans up on full success.

Run All UI Tests

Execute all UI test markers sequentially, recording failures and cleaning up on success.

UI Test Markers (from pytest.ini)

Run in this order:

  1. •splash_ui
  2. •home_ui
  3. •utubs_ui
  4. •members_ui
  5. •urls_ui
  6. •create_urls_ui
  7. •update_urls_ui
  8. •tags_ui
  9. •mobile_ui

Procedure

1. Setup

Create tmp directory and set failure file path:

bash
mkdir -p tmp
FAILURE_FILE="tmp/UI_$(date +%Y%m%d_%H%M%S)_failures.txt"

2. Run Each Marker Sequentially

For each marker above, run:

bash
docker exec u4i-local-web /bin/bash -c "source /code/venv/bin/activate && pytest -m 'MARKER'"
  • •Wait for each test suite to complete before starting the next
  • •If tests fail, append marker name and failure summary to the failure file
  • •Continue to the next marker regardless of pass/fail

3. On Completion

If any failures occurred:

  • •Report which markers failed
  • •Reference the failure file: tmp/UI_<timestamp>_failures.txt
  • •MUST include a snippet of the stack trace indicating why the error occurred

If ALL tests passed:

  • •Remove all UI failure files:
bash
rm -f tmp/UI_*
  • •Confirm success and cleanup