Report Generation

When to Use

Controlling where reports land and how they relate to test runs.

Default Location

playwright-report/ at the project root, written at the end of the run.

Pattern: Override Location

reporter: [['html', { outputFolder: 'reports/html' }]],

Per-Run Behavior

Playwright overwrites the contents of outputFolder at the start of every run. To keep historical reports:

Approach	When
Configure unique outputFolder per run (timestamped)	Local long-term archive
Copy/zip playwright-report/ after the run	Simple, manual
Use the blob reporter (['blob']), then npx playwright merge-reports --reporter=html	Sharded CI runs

Decision: playwright-report/ vs test-results/

Folder	Contains
playwright-report/	The HTML SPA + bundled attachments — the triage UI
test-results/	Raw per-test attachments (failure screenshots, traces, videos) — referenced by the report at generation; safe to delete after report is built

Both are gitignored by convention. Only commit *-snapshots/ (the baselines next to test files).

Pattern: Blob Reporter for Sharded Runs

// On each shard
reporter: [['blob']]

# After all shards
npx playwright merge-reports --reporter=html ./all-blobs

Produces one unified HTML report from N parallel shards.

Common Mistakes

• Committing playwright-report/ — clutters the repo; regeneratable
• Expecting reports to accumulate without explicit copying — they don't; every run overwrites
• Confusing outputFolder (HTML report) with outputDir (test-results) — different config keys, different folders

See Also

• Viewing
• Sharing
• Configuration Reference