Exact request sent to the run panel.
{
"source": "retrieve",
"template": {
"id": "extract_social_posts",
"slug": "extract-posts",
"name": "Extract Social Posts",
"categoryId": "EXTRACT"
},
"inputs": {
"platform": "auto",
"target": "",
"postLimit": 20,
"includeMetrics": true,
"analyzeImages": true,
"normalizedTarget": {
"raw": ""
}
},
"agentHint": "You are extracting social posts with optional image analysis. Adapt to the platform.\n\n**TASK CONFIGURATION:**\n- Platform: auto\n- Target: \n- Post Limit: 20\n- Include Metrics: true\n- Analyze Images: true\n\n---\n\n## STEP 1: NAVIGATE TO PROFILE\n\n1. Parse target - detect platform if \"auto\"\n2. Navigate to profile\n3. Wait for full load\n4. **Verify access:**\n - Profile exists? If not, report error\n - Private? Report limitation\n - Posts visible? Proceed\n\n---\n\n## STEP 2: EXTRACT PROFILE DATA\n\nCapture:\n- Handle/username\n- Display name\n- Bio text\n- Profile picture URL\n- Follower count\n- Following count\n- Post count\n- Verification status\n- External links\n\n---\n\n## STEP 3: EXTRACT POSTS\n\n**For each post (up to 20):**\n\n### Basic Data\n- Post URL/permalink\n- Post date/timestamp\n- Post type (image, carousel, video, text, reel, story highlight)\n\n### Text Content\n- Full caption/text\n- Hashtags (extracted separately)\n- Mentions (extracted separately)\n- Links included\n\n### Media Content\n- **Image URLs** (all images for carousels)\n- **Video thumbnail URL** (for videos)\n- **Media count** (for carousels)\n\n### Engagement (if true)\n- Likes/reactions count\n- Comments count\n- Shares/reposts count\n- Views (for videos if available)\n- Saves (if visible)\n\n---\n\n## STEP 4: IMAGE ANALYSIS (if true)\n\n**For each post with images, generate detailed descriptions:**\n\n### For Single Images:\nDescribe:\n- **Subject**: What is the primary subject?\n- **People**: Are there people? Expressions? Poses?\n- **Setting**: Where was this taken? Indoor/outdoor?\n- **Objects**: Notable objects or products?\n- **Colors**: Dominant color palette\n- **Style**: Photography style, filter, editing\n- **Text**: Any text overlay? What does it say?\n- **Quality**: Professional vs. casual\n- **Category**: Classify (selfie, product, landscape, graphic, etc.)\n\n### For Carousels:\nDescribe each slide:\n- Slide-by-slide descriptions\n- Overall narrative/theme\n- Hook slide analysis\n- CTA slide analysis\n\n### For Videos:\nDescribe thumbnail:\n- What the preview frame shows\n- Hook elements visible\n- Text overlays\n\n**Take appropriate time for thorough analysis.**\n\n---\n\n## STEP 5: CATEGORIZE VISUAL CONTENT\n\n**Across all posts, categorize:**\n\nVisual content types found:\n- Selfies/portraits\n- Product shots\n- Lifestyle photos\n- Graphics/infographics\n- Screenshots\n- User-generated content\n- Behind-the-scenes\n- Text-on-image posts\n- Memes\n\n---\n\n## STEP 6: REPORT STRUCTURED DATA\n\n```json\n{\n \"profile\": {\n \"platform\": \"\",\n \"handle\": \"\",\n \"name\": \"\",\n \"bio\": \"\",\n \"profileUrl\": \"\",\n \"profilePicUrl\": \"\",\n \"profilePicDescription\": \"\",\n \"followers\": <number>,\n \"following\": <number>,\n \"postCount\": <number>,\n \"verified\": false,\n \"externalLinks\": []\n },\n \"extraction\": {\n \"postsRequested\": 20,\n \"postsExtracted\": <number>,\n \"postsWithImages\": <number>,\n \"imagesAnalyzed\": <number>,\n \"extractedAt\": \"<ISO timestamp>\"\n },\n \"posts\": [\n {\n \"postUrl\": \"\",\n \"date\": \"<ISO date>\",\n \"type\": \"image|carousel|video|text|reel\",\n \"caption\": \"\",\n \"hashtags\": [],\n \"mentions\": [],\n \"media\": {\n \"type\": \"image|carousel|video|none\",\n \"urls\": [],\n \"count\": <number>\n },\n \"imageAnalysis\": [\n {\n \"imageUrl\": \"\",\n \"description\": \"detailed description\",\n \"subjects\": [],\n \"setting\": \"\",\n \"colors\": [],\n \"style\": \"\",\n \"textOverlay\": \"\",\n \"category\": \"\",\n \"quality\": \"\"\n }\n ],\n \"metrics\": {\n \"likes\": <number>,\n \"comments\": <number>,\n \"shares\": <number>,\n \"views\": <number or null>,\n \"saves\": <number or null>\n }\n }\n ],\n \"visualContentSummary\": {\n \"totalImages\": <number>,\n \"contentCategories\": {\n \"selfies\": <count>,\n \"products\": <count>,\n \"lifestyle\": <count>,\n \"graphics\": <count>,\n \"screenshots\": <count>,\n \"other\": <count>\n },\n \"dominantColors\": [],\n \"photographyStyle\": \"\",\n \"averageQuality\": \"\"\n }\n}\n```\n\n---\n\n## IMAGE ANALYSIS DEPTH\n\n**With true = true:**\n- Each image gets detailed description\n- Descriptions are structured and consistent\n- Categories are assigned\n- Visual patterns are summarized\n\n**This adds processing time but provides rich data for:**\n- Content strategy analysis\n- Aesthetic assessment\n- Competitive intelligence\n- Training data extraction",
"userInput": "You are extracting social posts with optional image analysis. Adapt to the platform.\n\n**TASK CONFIGURATION:**\n- Platform: auto\n- Target: \n- Post Limit: 20\n- Include Metrics: true\n- Analyze Images: true\n\n---\n\n## STEP 1: NAVIGATE TO PROFILE\n\n1. Parse target - detect platform if \"auto\"\n2. Navigate to profile\n3. Wait for full load\n4. **Verify access:**\n - Profile exists? If not, report error\n - Private? Report limitation\n - Posts visible? Proceed\n\n---\n\n## STEP 2: EXTRACT PROFILE DATA\n\nCapture:\n- Handle/username\n- Display name\n- Bio text\n- Profile picture URL\n- Follower count\n- Following count\n- Post count\n- Verification status\n- External links\n\n---\n\n## STEP 3: EXTRACT POSTS\n\n**For each post (up to 20):**\n\n### Basic Data\n- Post URL/permalink\n- Post date/timestamp\n- Post type (image, carousel, video, text, reel, story highlight)\n\n### Text Content\n- Full caption/text\n- Hashtags (extracted separately)\n- Mentions (extracted separately)\n- Links included\n\n### Media Content\n- **Image URLs** (all images for carousels)\n- **Video thumbnail URL** (for videos)\n- **Media count** (for carousels)\n\n### Engagement (if true)\n- Likes/reactions count\n- Comments count\n- Shares/reposts count\n- Views (for videos if available)\n- Saves (if visible)\n\n---\n\n## STEP 4: IMAGE ANALYSIS (if true)\n\n**For each post with images, generate detailed descriptions:**\n\n### For Single Images:\nDescribe:\n- **Subject**: What is the primary subject?\n- **People**: Are there people? Expressions? Poses?\n- **Setting**: Where was this taken? Indoor/outdoor?\n- **Objects**: Notable objects or products?\n- **Colors**: Dominant color palette\n- **Style**: Photography style, filter, editing\n- **Text**: Any text overlay? What does it say?\n- **Quality**: Professional vs. casual\n- **Category**: Classify (selfie, product, landscape, graphic, etc.)\n\n### For Carousels:\nDescribe each slide:\n- Slide-by-slide descriptions\n- Overall narrative/theme\n- Hook slide analysis\n- CTA slide analysis\n\n### For Videos:\nDescribe thumbnail:\n- What the preview frame shows\n- Hook elements visible\n- Text overlays\n\n**Take appropriate time for thorough analysis.**\n\n---\n\n## STEP 5: CATEGORIZE VISUAL CONTENT\n\n**Across all posts, categorize:**\n\nVisual content types found:\n- Selfies/portraits\n- Product shots\n- Lifestyle photos\n- Graphics/infographics\n- Screenshots\n- User-generated content\n- Behind-the-scenes\n- Text-on-image posts\n- Memes\n\n---\n\n## STEP 6: REPORT STRUCTURED DATA\n\n```json\n{\n \"profile\": {\n \"platform\": \"\",\n \"handle\": \"\",\n \"name\": \"\",\n \"bio\": \"\",\n \"profileUrl\": \"\",\n \"profilePicUrl\": \"\",\n \"profilePicDescription\": \"\",\n \"followers\": <number>,\n \"following\": <number>,\n \"postCount\": <number>,\n \"verified\": false,\n \"externalLinks\": []\n },\n \"extraction\": {\n \"postsRequested\": 20,\n \"postsExtracted\": <number>,\n \"postsWithImages\": <number>,\n \"imagesAnalyzed\": <number>,\n \"extractedAt\": \"<ISO timestamp>\"\n },\n \"posts\": [\n {\n \"postUrl\": \"\",\n \"date\": \"<ISO date>\",\n \"type\": \"image|carousel|video|text|reel\",\n \"caption\": \"\",\n \"hashtags\": [],\n \"mentions\": [],\n \"media\": {\n \"type\": \"image|carousel|video|none\",\n \"urls\": [],\n \"count\": <number>\n },\n \"imageAnalysis\": [\n {\n \"imageUrl\": \"\",\n \"description\": \"detailed description\",\n \"subjects\": [],\n \"setting\": \"\",\n \"colors\": [],\n \"style\": \"\",\n \"textOverlay\": \"\",\n \"category\": \"\",\n \"quality\": \"\"\n }\n ],\n \"metrics\": {\n \"likes\": <number>,\n \"comments\": <number>,\n \"shares\": <number>,\n \"views\": <number or null>,\n \"saves\": <number or null>\n }\n }\n ],\n \"visualContentSummary\": {\n \"totalImages\": <number>,\n \"contentCategories\": {\n \"selfies\": <count>,\n \"products\": <count>,\n \"lifestyle\": <count>,\n \"graphics\": <count>,\n \"screenshots\": <count>,\n \"other\": <count>\n },\n \"dominantColors\": [],\n \"photographyStyle\": \"\",\n \"averageQuality\": \"\"\n }\n}\n```\n\n---\n\n## IMAGE ANALYSIS DEPTH\n\n**With true = true:**\n- Each image gets detailed description\n- Descriptions are structured and consistent\n- Categories are assigned\n- Visual patterns are summarized\n\n**This adds processing time but provides rich data for:**\n- Content strategy analysis\n- Aesthetic assessment\n- Competitive intelligence\n- Training data extraction",
"urls": [],
"meta": {
"createdAt": "",
"userAgent": ""
},
"execution": {
"mode": "device",
"deviceIds": []
}
}
Add to Chrome