generating-test-reports

SKILL.md

Test Report Generator

Overview

Generate structured, human-readable test reports from raw test runner output including JUnit XML, Jest JSON, pytest results, and coverage data. Produces Markdown summaries, HTML dashboards, and CI-compatible annotations.

Prerequisites

  • Test results in a parseable format (JUnit XML, Jest --json, pytest --junitxml, or TAP)
  • Coverage data files (Istanbul coverage-summary.json, lcov.info, or coverage.xml)
  • Node.js or Python available for report generation scripts
  • Git history accessible for trend analysis across commits

Instructions

  1. Locate all test result files using Glob patterns (**/junit.xml, **/test-results.json, **/coverage/lcov.info).
  2. Parse each result file and extract:
    • Total test count, passed, failed, skipped, and error counts.
    • Execution duration per test suite and per individual test.
    • Failure messages, stack traces, and assertion details.
  3. Parse coverage data and extract:
    • Line, branch, function, and statement coverage percentages.
    • Per-file coverage breakdown identifying files below threshold.
    • Uncovered line ranges for targeted improvement.
  4. Compute aggregate metrics:
    • Overall pass rate as a percentage.
    • Total execution time and average test duration.
    • Top 10 slowest tests with file paths and durations.
    • Coverage delta compared to the previous commit (if git history is available).
  5. Generate a Markdown report with sections for summary, failures, coverage, and performance.
  6. Optionally generate an HTML report with sortable tables and coverage heatmaps.
  7. Write CI-compatible output (GitHub Actions job summary, GitLab report artifacts, or Slack webhook payload).

Output

  • test-report.md -- Markdown summary with pass/fail table, coverage stats, and failure details
  • test-report.html -- Self-contained HTML report (optional)
  • Coverage summary table with per-file breakdown and delta from baseline
  • Slowest tests list ranked by execution time
  • CI annotation comments on failed test lines (GitHub Actions ::error format)

Error Handling

Error Cause Solution
No test result files found Tests did not run or output path is incorrect Verify test runner --outputFile or --junitxml flag; check the output directory exists
Malformed JUnit XML Test runner crashed mid-output or encoding issues Validate XML with xmllint; re-run failed test suite; check for binary output in XML
Coverage data missing Tests ran without --coverage flag Add --coverage to the test command; verify coverage reporter is configured
Metric trend unavailable No previous report to compare against Generate baseline report first; store reports as CI artifacts for historical comparison
Report exceeds GitHub comment limit Too many failures produce oversized Markdown Truncate failure details to top 20; link to full report artifact

Examples

Markdown report structure:

## Test Results -- 2026-03-10

| Metric | Value |
|--------|-------|
| Total Tests | 847 |  # 847 = configured value
| Passed | 839 (99.1%) |  # 839 = configured value
| Failed | 5 |
| Skipped | 3 |
| Duration | 42.3s |

### Coverage
| Category | Current | Threshold | Status |
|----------|---------|-----------|--------|
| Lines | 87.2% | 80% | PASS |
| Branches | 74.1% | 70% | PASS |
| Functions | 91.5% | 85% | PASS |

### Failed Tests
1. `src/utils/parser.test.ts` -- "handles malformed input" -- Expected Error but received null
2. `src/api/auth.test.ts` -- "rejects expired tokens" -- Timeout after 5000ms

GitHub Actions job summary integration:

cat test-report.md >> "$GITHUB_STEP_SUMMARY"

Resources

Weekly Installs
21
GitHub Stars
1.6K
First Seen
Feb 4, 2026
Installed on
codex20
gemini-cli19
claude-code19
github-copilot19
kimi-cli19
amp19