You are analyzing web agency client feedback from a feedback tool called Feedbucket. Some entries are from trial users just testing the widget.

Classify each feedback entry and determine if it requires client-specific knowledge.

CATEGORIES:
1: Visual Design - Photos, images, colors, fonts, styling, layout, animations, logos, visual appearance
2: Content & Copy - Text updates, titles, copy changes, spelling corrections, translations, contact info
3: Functionality/Bugs - Things not working, broken features, errors, malfunctions, display issues
4: User Experience (UX) - Navigation, usability, user flow improvements, ease of use, interaction design
5: Performance - Speed, loading times, mobile responsiveness, optimization issues
6: Feature Requests - Adding new functionality, missing features, new integrations, enhancements
7: Technical/Development - Code changes, SEO, accessibility, browser compatibility, backend work
8: Project Management - Timeline discussions, communication, process feedback, project coordination
9: Test Entry - Trial users testing the feedback widget with dummy/test content
10: Uncategorized - Feedback that doesn't clearly fit into the above categories

IMPORTANT: Look for test entries from trial users:
- Generic test content like "test", "testing", "hello", "demo"
- Obviously fake feedback like "this is a test"
- Placeholder content or very generic entries
- Content that doesn't relate to actual website feedback

For each feedback entry, provide:
1. The category number (1-10)
2. A brief reason explaining your classification
3. Whether it's client-specific (true/false)

REASONING LENGTH:
- Categories 1-9: Keep reasoning concise (max 12 words)
- Category 10 (Uncategorized): Provide detailed reasoning (20-40 words) explaining what the feedback is about and why it doesn't fit other categories

NOTE: These are just examples to help you understand the categories:
- "curve disappears when background changes" would be Category 3, client_specific: false (technical bug)
- "cambiar foto" would be Category 1, client_specific: true (client wants specific photo)
- "añadir WhatsApp button" would be Category 6, client_specific: true (client-specific integration)
- "button not working" would be Category 3, client_specific: false (objective technical issue)
- "make text bigger" would be Category 1, client_specific: true (aesthetic preference)

When unsure between categories, choose the most specific one. Only use Category 10 (Uncategorized) when the feedback truly doesn't fit any of the other 9 categories.

Feedback to classify:
{feedback_text}

Respond in this exact JSON format:
{{
  "classifications": [
    {{"id": "370880", "category": 3, "reasoning": "Visual element breaks under certain conditions", "client_specific": false}},
    {{"id": "370878", "category": 1, "reasoning": "Logo visibility issue on dark background", "client_specific": true}}
  ]
}}

Only return the JSON, no other text.