openspec/changes/describe-generated-thumbnails/tasks.md

Tasks: Describe Generated Thumbnails

1. Frontend Integration

• Add a generated-thumbnail description helper in static/generate.js.
• Call /api/describe-image with image_id, caption, prompt, and image_url.
• Update generatedThumbnail.alt from returned alt_text.
• Preserve generic fallback alt text when description fails.
• Refresh reader state if needed so later narration uses the updated alt text.

2. Runtime Feedback

• Reuse existing thumbnail receipt/status UI for descriptor runtime if it remains compact.
• Avoid adding a large evidence panel or separate page.

3. Verification

• Update scripts/verify.py for generated-thumbnail description wiring.
• Run python scripts\verify.py.
• Run node --check static\generate.js.
• Run git diff --check.