AgentSkillsCN

test-failure-logger

在开发、测试或 CI/CD 流程中，每遇到一次测试失败，都应自动将失败信息记录至 tmp/ 目录下的带时间戳文件中。该技能会捕获测试名称、时间戳、可能的原因分析，以及堆栈跟踪片段。对于追踪功能开发与调试过程中的测试失败问题至关重要。

SKILL.md
--- frontmatter
name: test-failure-logger
description: Automatically log test failures to timestamped files in tmp/ directory. Use this skill EVERY TIME a test failure is encountered during development, testing, or CI/CD workflows. Captures test name, timestamp, likely cause analysis, and stack trace snippets. Critical for tracking test failures across feature development and debugging sessions.

Test Failure Logger

Automatically capture and log test failure information whenever tests fail.

When to Use

CRITICAL: Use this skill EVERY TIME you encounter a test failure, regardless of other tasks. Test failure tracking is essential to feature development and debugging.

Triggers include:

  • •pytest failures
  • •UI/functional test failures (Selenium, Playwright)
  • •Integration test failures
  • •Any test suite execution that encounters failures

Workflow

When a test fails:

  1. •

    Extract failure information:

    • •Test name/path (e.g., tests/functional/splash_ui/test_reset_password_ui.py::test_reset_password_success)
    • •Likely cause (brief analysis: "Missing element on page", "API returned 500", "Database connection timeout")
    • •Stack trace snippet (the relevant error lines, not the full trace)
  2. •

    Log the failure:

    python
    python .claude/skills/test-failure-logger/scripts/log_test_failure.py \
  "test_name" \
  "likely_cause" \
  "stack_trace_snippet"
  3. •

    Continue with your task - The failure is now logged for future reference

Output

Creates timestamped JSON files in tmp/ directory:

  • •Filename pattern: test_failure_YYYYMMDD_HHMMSS_<sanitized_test_name>.json
  • •Contains: test name, ISO timestamp, likely cause, stack trace, optional context

Example

bash
python .claude/skills/test-failure-logger/scripts/log_test_failure.py \
  "tests/functional/splash_ui/test_login.py::test_login_invalid_credentials" \
  "Assertion failed - expected error message not displayed" \
  "AssertionError: assert False where False = <element>.is_displayed()"

Integration with Workflows

  • •During active development: Log failures immediately when encountered
  • •After test runs: Log all failures from the test output
  • •CI/CD failures: Log the specific test that caused the build to fail
  • •Debugging sessions: Log failures as you investigate them

Cleanup

Logged failures in tmp/ are temporary and can be cleared when:

  • •The issue is resolved and verified
  • •Starting a fresh development cycle
  • •The information is no longer relevant

Run rm tmp/test_failure_*.json to clear all logged failures.