Runner Console

The Runner Console is a live-streaming terminal embedded in the Runner panel. It renders Playwright's stdout and stderr with intelligent formatting, encoding protection, and integrated AI analysis.

Live Streaming

Output appears line-by-line as Playwright emits it — there is no buffering delay. The console auto-scrolls to the latest output during execution.

Running 47 tests using 4 workers

  [pass] login.spec.ts:12 — should accept valid credentials (1.2s)
  [pass] login.spec.ts:28 — should reject wrong password (0.8s)
  [FAIL] cart.spec.ts:45  — should update quantity (3.1s)
  [pass] checkout.spec.ts:15 — should calculate tax (1.0s)

Pause Scroll

Click anywhere in the console to pause auto-scroll. A Resume button appears at the bottom to jump back to the latest output.

Mojibake Protection

Playwright child processes sometimes emit output with mismatched encodings — especially on Windows with non-ASCII characters. The console detects and corrects this automatically:

Problem	Cause	Fix Applied
Garbled umlauts	UTF-8 bytes read as Latin-1	Re-decode as UTF-8
??? replacement chars	Missing code page	Fallback to system locale
ANSI escape fragments	Partial SGR sequences	Strip incomplete sequences

This runs transparently — you see clean, readable text regardless of your system locale or Playwright configuration.

Error Highlighting

Failed tests render with distinct visual treatment:

• Red background on the failure summary line
• Stack trace with syntax-highlighted file paths
• Source link — click a file path (e.g. cart.spec.ts:45:12) to open it in the editor at the exact line
• Expected vs. Received — assertion diffs render in a side-by-side comparison

  [FAIL] cart.spec.ts:45 — should update quantity

    Error: expect(received).toBe(expected)
    Expected: 3
    Received: 2

    at tests/cart.spec.ts:45:12

AI Analysis

Each failed test displays a Brain icon button. Clicking it sends the error context — message, stack trace, surrounding source code — to the configured AI provider.

The AI returns:

1. Root cause — plain-language explanation of why the test failed
2. Suggested fix — code diff you can apply with one click
3. Confidence score — how certain the model is about the diagnosis

AI Provider

AI Analysis uses whichever provider is configured in Settings > AI > Provider (OpenAI, Anthropic, Google, or local proxy). The feature requires a valid API key.

Console Controls

Action	Shortcut	Description
Clear	Ctrl+L	Clears the console buffer
Copy All	Ctrl+Shift+C	Copies full console output to clipboard
Search	Ctrl+F	Searches within console output
Filter Errors	—	Toggle to show only failed test output

Buffer Limit

The console retains the last 50,000 lines. Older output is discarded to keep memory usage stable. For full logs, use the CI Integration to write JUnit XML or JSON reports.