[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M6XDHA916CXQ38FMHXDG5R84":3,"skill-same:01M6XDHA916CXQ38FMHXDG5R84":54},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":28,"linkOnly":29,"markdownContent":30,"markdownBody":31,"frontMatter":32,"refName":33,"commitSha":34,"skillDirectory":35,"forksCount":13,"repositoryHtmlUrl":36,"repositoryHomepageUrl":6,"ownerHtmlUrl":37,"ownerBlogUrl":6,"ownerDescription":38,"updatedAt":39,"install":40,"resources":53},"01M6XDHA916CXQ38FMHXDG5R84","gpt-multimodal",null,"Analyze images and multi-frame sequences using OpenAI GPT series","Waaaalnut\u002Fskillsbench","Waaaalnut","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F103995656?v=4","01M6XCS9K1V51EY3JMB7DGMX24","01M6XCS9Q4WPW88W62C55G1H44",0,"Apache-2.0","https:\u002F\u002Fgithub.com\u002FWaaaalnut\u002Fskillsbench\u002Fblob\u002Fmain\u002Ftasks-no-skills\u002Fpedestrian-traffic-counting\u002Fenvironment\u002Fskills\u002Fgpt-multimodal\u002FSKILL.md",[],[18,19,20,21,22,23,24,25,26,27],"multimodal","gpt","gpt multimodal","openai","analyze","images","series","multi","frame","sequences","2026-11-10 10:12:30",false,"---\nname: gpt-multimodal\ndescription: Analyze images and multi-frame sequences using OpenAI GPT series\n---\n\n# OpenAI Vision Analysis Skill\n\n## Purpose\nThis skill enables image analysis, scene understanding, text extraction, and multi-frame comparison using OpenAI's vision-capable GPT models (e.g., `gpt-4o`, `gpt-5`). It supports single and multiple images analysis and sequential frames for temporal analysis.\n\n## When to Use\n- Analyzing image content (objects, scenes, colors, spatial relationships)\n- Extracting and reading text from images (OCR via vision models)\n- Comparing multiple images to detect differences or changes\n- Processing video frames to understand temporal progression\n- Generating detailed image descriptions or captions\n- Answering questions about visual content\n\n## Required Libraries\n\nThe following Python libraries are required:\n\n```python\nfrom openai import OpenAI\nimport base64\nimport json\nimport os\nfrom pathlib import Path\n```\n\n## Input Requirements\n- **File formats**: JPG, JPEG, PNG, WEBP, non-animated GIF\n- **Image quality**: Clear and legible; minimum 512×512px recommended\n- **File size**: Under 20MB per image recommended\n- **Maximum per request**: Up to 500 images, 50MB total payload\n- **URL or Base64**: Images can be provided as URLs or base64-encoded data\n\n## Output Schema\nAll analysis results should be returned as valid JSON conforming to this schema:\n\n```json\n{\n  \"success\": true,\n  \"model\": \"gpt-5\",\n  \"analysis\": \"Detailed description or analysis of the image content...\",\n  \"metadata\": {\n    \"image_count\": 1,\n    \"detail_level\": \"high\",\n    \"tokens_used\": 850,\n    \"processing_time_ms\": 1234\n  },\n  \"extracted_data\": {\n    \"objects\": [\"car\", \"person\", \"building\"],\n    \"text_found\": \"Sample text from image\",\n    \"colors\": [\"blue\", \"white\", \"gray\"],\n    \"scene_type\": \"urban street\"\n  },\n  \"warnings\": []\n}\n```\n\n### Field Descriptions\n\n- `success`: Boolean indicating whether the API call succeeded\n- `model`: The GPT model used for analysis (e.g., \"gpt-4o\", \"gpt-5\")\n- `analysis`: Complete textual analysis or description from the model\n- `metadata.image_count`: Number of images analyzed in this request\n- `metadata.detail_level`: Detail parameter used (\"low\", \"high\", or \"auto\")\n- `metadata.tokens_used`: Approximate token count for the request\n- `metadata.processing_time_ms`: Time taken to process the request\n- `extracted_data`: Structured information extracted from the image(s)\n- `warnings`: Array of issues or limitations encountered\n\n\n## Code Examples\n\n### Basic Image Analysis\n\n```python\nfrom openai import OpenAI\nimport base64\n\ndef analyze_image(image_path, prompt=\"What's in this image?\"):\n    \"\"\"Analyze a single image using GPT-5 Vision.\"\"\"\n    client = OpenAI(api_key=os.environ.get(\"OPENAI_API_KEY\"))\n    \n    # Read and encode image\n    with open(image_path, \"rb\") as image_file:\n        base64_image = base64.b64encode(image_file.read()).decode('utf-8')\n    \n    response = client.chat.completions.create(\n        model=\"gpt-5\",\n        messages=[\n            {\n                \"role\": \"user\",\n                \"content\": [\n                    {\"type\": \"text\", \"text\": prompt},\n                    {\n                        \"type\": \"image_url\",\n                        \"image_url\": {\n                            \"url\": f\"data:image\u002Fjpeg;base64,{base64_image}\"\n                        }\n                    }\n                ]\n            }\n        ],\n        max_tokens=300\n    )\n    \n    return response.choices[0].message.content\n```\n\n### Using Image URLs\n\n```python\nfrom openai import OpenAI\n\ndef analyze_image_url(image_url, prompt=\"Describe this image\"):\n    \"\"\"Analyze an image from a URL.\"\"\"\n    client = OpenAI(api_key=os.environ.get(\"OPENAI_API_KEY\"))\n    \n    response = client.chat.completions.create(\n        model=\"gpt-5\",\n        messages=[\n            {\n                \"role\": \"user\",\n                \"content\": [\n                    {\"type\": \"text\", \"text\": prompt},\n                    {\n                        \"type\": \"image_url\",\n                        \"image_url\": {\"url\": image_url}\n                    }\n                ]\n            }\n        ]\n    )\n    \n    return response.choices[0].message.content\n```\n\n### Multiple Images Analysis\n\n```python\nfrom openai import OpenAI\nimport base64\n\ndef analyze_multiple_images(image_paths, prompt=\"Compare these images\"):\n    \"\"\"Analyze multiple images in a single request.\"\"\"\n    client = OpenAI(api_key=os.environ.get(\"OPENAI_API_KEY\"))\n    \n    # Build content array with text and all images\n    content = [{\"type\": \"text\", \"text\": prompt}]\n    \n    for image_path in image_paths:\n        with open(image_path, \"rb\") as image_file:\n            base64_image = base64.b64encode(image_file.read()).decode('utf-8')\n        \n        content.append({\n            \"type\": \"image_url\",\n            \"image_url\": {\n                \"url\": f\"data:image\u002Fjpeg;base64,{base64_image}\"\n            }\n        })\n    \n    response = client.chat.completions.create(\n        model=\"gpt-5\",\n        messages=[{\"role\": \"user\", \"content\": content}],\n        max_tokens=500\n    )\n    \n    return response.choices[0].message.content\n```\n\n### Full Analysis with JSON Output\n\n```python\nfrom openai import OpenAI\nimport base64\nimport json\nimport time\n\ndef analyze_image_to_json(image_path, prompt=\"Analyze this image\"):\n    \"\"\"Analyze image and return structured JSON output.\"\"\"\n    client = OpenAI(api_key=os.environ.get(\"OPENAI_API_KEY\"))\n    \n    start_time = time.time()\n    warnings = []\n    \n    try:\n        # Read and encode image\n        with open(image_path, \"rb\") as image_file:\n            base64_image = base64.b64encode(image_file.read()).decode('utf-8')\n        \n        # Make API call\n        response = client.chat.completions.create(\n            model=\"gpt-5\",\n            messages=[\n                {\n                    \"role\": \"user\",\n                    \"content\": [\n                        {\"type\": \"text\", \"text\": prompt},\n                        {\n                            \"type\": \"image_url\",\n                            \"image_url\": {\n                                \"url\": f\"data:image\u002Fjpeg;base64,{base64_image}\",\n                                \"detail\": \"high\"\n                            }\n                        }\n                    ]\n                }\n            ],\n            max_tokens=500\n        )\n        \n        analysis = response.choices[0].message.content\n        tokens_used = response.usage.total_tokens\n        processing_time = int((time.time() - start_time) * 1000)\n        \n        result = {\n            \"success\": True,\n            \"model\": \"gpt-5\",\n            \"analysis\": analysis,\n            \"metadata\": {\n                \"image_count\": 1,\n                \"detail_level\": \"high\",\n                \"tokens_used\": tokens_used,\n                \"processing_time_ms\": processing_time\n            },\n            \"extracted_data\": {},\n            \"warnings\": warnings\n        }\n        \n    except Exception as e:\n        result = {\n            \"success\": False,\n            \"model\": \"gpt-5\",\n            \"analysis\": \"\",\n            \"metadata\": {\n                \"image_count\": 0,\n                \"detail_level\": \"high\",\n                \"tokens_used\": 0,\n                \"processing_time_ms\": 0\n            },\n            \"extracted_data\": {},\n            \"warnings\": [f\"API call failed: {str(e)}\"]\n        }\n    \n    return result\n\n# Usage\nresult = analyze_image_to_json(\"photo.jpg\", \"Describe what you see in detail\")\nprint(json.dumps(result, indent=2))\n```\n\n### Batch Processing with Sequential Frames\n\n```python\nfrom openai import OpenAI\nimport base64\nfrom pathlib import Path\n\ndef process_video_frames(frames_directory, analysis_prompt):\n    \"\"\"Process sequential video frames for temporal analysis.\"\"\"\n    client = OpenAI(api_key=os.environ.get(\"OPENAI_API_KEY\"))\n    \n    image_extensions = {'.jpg', '.jpeg', '.png', '.webp'}\n    frame_paths = sorted([\n        f for f in Path(frames_directory).iterdir() \n        if f.suffix.lower() in image_extensions\n    ])\n    \n    # Analyze frames in groups (e.g., 5 frames at a time)\n    batch_size = 5\n    results = []\n    \n    for i in range(0, len(frame_paths), batch_size):\n        batch = frame_paths[i:i+batch_size]\n        \n        # Build content with all frames in batch\n        content = [{\"type\": \"text\", \"text\": analysis_prompt}]\n        \n        for frame_path in batch:\n            with open(frame_path, \"rb\") as f:\n                base64_image = base64.b64encode(f.read()).decode('utf-8')\n            \n            content.append({\n                \"type\": \"image_url\",\n                \"image_url\": {\n                    \"url\": f\"data:image\u002Fjpeg;base64,{base64_image}\",\n                    \"detail\": \"low\"  # Use low detail for video frames to save tokens\n                }\n            })\n        \n        response = client.chat.completions.create(\n            model=\"gpt-5\",\n            messages=[{\"role\": \"user\", \"content\": content}],\n            max_tokens=800\n        )\n        \n        results.append({\n            \"batch_index\": i \u002F\u002F batch_size,\n            \"frame_range\": f\"{batch[0].name} to {batch[-1].name}\",\n            \"analysis\": response.choices[0].message.content\n        })\n    \n    return results\n```\n\n### Text Extraction from Images (OCR Alternative)\n\n```python\nfrom openai import OpenAI\nimport base64\n\ndef extract_text_with_gpt(image_path):\n    \"\"\"Extract text from image using GPT Vision as OCR alternative.\"\"\"\n    client = OpenAI(api_key=os.environ.get(\"OPENAI_API_KEY\"))\n    \n    with open(image_path, \"rb\") as image_file:\n        base64_image = base64.b64encode(image_file.read()).decode('utf-8')\n    \n    response = client.chat.completions.create(\n        model=\"gpt-5\",\n        messages=[\n            {\n                \"role\": \"user\",\n                \"content\": [\n                    {\n                        \"type\": \"text\", \n                        \"text\": \"Extract all text from this image. Return only the text content, preserving the layout and structure.\"\n                    },\n                    {\n                        \"type\": \"image_url\",\n                        \"image_url\": {\n                            \"url\": f\"data:image\u002Fjpeg;base64,{base64_image}\",\n                            \"detail\": \"high\"\n                        }\n                    }\n                ]\n            }\n        ],\n        max_tokens=1000\n    )\n    \n    return response.choices[0].message.content\n```\n\n\n## Model Selection and Configuration\n\n### Available Models\n\n```python\n# GPT-4o - Best for general vision tasks, fast and cost-effective\nmodel = \"gpt-4o\"\n\n# GPT-5-nano - Faster and cheaper for simple vision tasks\nmodel = \"gpt-5-nano\"\n\n# GPT-5 - More capable for complex reasoning\nmodel = \"gpt-5\"\n```\n\n### Detail Level Configuration\n\nControl how much visual detail the model processes:\n\n```python\n# Low detail - 512×512px resolution, fewer tokens, faster\n\"image_url\": {\n    \"url\": image_url,\n    \"detail\": \"low\"\n}\n\n# High detail - Full resolution with tiling, more tokens, better accuracy\n\"image_url\": {\n    \"url\": image_url,\n    \"detail\": \"high\"\n}\n\n# Auto - Model chooses appropriate detail level\n\"image_url\": {\n    \"url\": image_url,\n    \"detail\": \"auto\"\n}\n```\n\n**When to use each detail level:**\n- **Low**: Video frames, simple scene classification, color\u002Fshape detection\n- **High**: Text extraction, detailed object detection, fine-grained analysis\n- **Auto**: General purpose when unsure; model optimizes cost vs. quality\n\n\n## Token Cost Management\n\n### Understanding Image Tokens\n\nImage tokens count toward your request limits and costs:\n\n- **Low detail**: Fixed ~85 tokens per image (gpt-5)\n- **High detail**: Base tokens + tile tokens based on image dimensions\n  - Images are scaled to fit within 2048×2048px\n  - Divided into 512×512px tiles\n  - Each tile costs additional tokens\n\n### Cost Calculation Examples\n\n```python\n# For gpt-5 with high detail:\n# - Base: 85 tokens\n# - Per tile: 170 tokens\n# - Example: 1024×1024 image = 85 + (2×2 tiles × 170) = 765 tokens\n\ndef estimate_tokens_high_detail(width, height):\n    \"\"\"Estimate token cost for high-detail image (gpt-5).\"\"\"\n    # Scale to fit within 2048×2048\n    scale = min(2048 \u002F width, 2048 \u002F height, 1.0)\n    scaled_w = int(width * scale)\n    scaled_h = int(height * scale)\n    \n    # Calculate tiles (512×512)\n    tiles_x = (scaled_w + 511) \u002F\u002F 512\n    tiles_y = (scaled_h + 511) \u002F\u002F 512\n    total_tiles = tiles_x * tiles_y\n    \n    # Token calculation\n    base_tokens = 85\n    tile_tokens = total_tiles * 170\n    \n    return base_tokens + tile_tokens\n```\n\n### Cost Optimization Strategies\n\n1. **Use low detail for video frames** - Temporal analysis doesn't need high resolution\n2. **Resize large images before uploading** - Reduce dimensions to 1024×1024 if high detail not needed\n3. **Batch related questions** - Analyze multiple aspects in one API call\n4. **Cache analysis results** - Store results for repeated processing\n\n\n## Advanced Use Cases\n\n### Image Comparison\n\n```python\ndef compare_images(image1_path, image2_path):\n    \"\"\"Compare two images and identify differences.\"\"\"\n    client = OpenAI(api_key=os.environ.get(\"OPENAI_API_KEY\"))\n    \n    images = []\n    for path in [image1_path, image2_path]:\n        with open(path, \"rb\") as f:\n            base64_image = base64.b64encode(f.read()).decode('utf-8')\n            images.append(base64_image)\n    \n    response = client.chat.completions.create(\n        model=\"gpt-5\",\n        messages=[\n            {\n                \"role\": \"user\",\n                \"content\": [\n                    {\n                        \"type\": \"text\",\n                        \"text\": \"Compare these two images. List all differences you observe, including changes in objects, colors, positions, or any other visual elements.\"\n                    },\n                    {\"type\": \"image_url\", \"image_url\": {\"url\": f\"data:image\u002Fjpeg;base64,{images[0]}\"}},\n                    {\"type\": \"image_url\", \"image_url\": {\"url\": f\"data:image\u002Fjpeg;base64,{images[1]}\"}}\n                ]\n            }\n        ]\n    )\n    \n    return response.choices[0].message.content\n```\n\n### Structured Data Extraction\n\n```python\ndef extract_structured_data(image_path, schema_description):\n    \"\"\"Extract structured information from image based on schema.\"\"\"\n    client = OpenAI(api_key=os.environ.get(\"OPENAI_API_KEY\"))\n    \n    with open(image_path, \"rb\") as f:\n        base64_image = base64.b64encode(f.read()).decode('utf-8')\n    \n    prompt = f\"\"\"Analyze this image and extract information in JSON format following this schema:\n{schema_description}\n\nReturn only valid JSON, no additional text.\"\"\"\n    \n    response = client.chat.completions.create(\n        model=\"gpt-5\",\n        messages=[\n            {\n                \"role\": \"user\",\n                \"content\": [\n                    {\"type\": \"text\", \"text\": prompt},\n                    {\n                        \"type\": \"image_url\",\n                        \"image_url\": {\n                            \"url\": f\"data:image\u002Fjpeg;base64,{base64_image}\",\n                            \"detail\": \"high\"\n                        }\n                    }\n                ]\n            }\n        ],\n        response_format={\"type\": \"json_object\"}\n    )\n    \n    return json.loads(response.choices[0].message.content)\n\n# Usage example\nschema = \"\"\"\n{\n  \"products\": [{\"name\": string, \"price\": number, \"quantity\": number}],\n  \"total\": number,\n  \"date\": string\n}\n\"\"\"\ndata = extract_structured_data(\"receipt.jpg\", schema)\n```\n\n\n## Error Handling\n\n### Common Issues and Solutions\n\n**Issue**: API authentication failed\n\n```python\n# Verify API key is set\nimport os\napi_key = os.environ.get(\"OPENAI_API_KEY\")\nif not api_key:\n    raise ValueError(\"OPENAI_API_KEY environment variable not set\")\n```\n\n**Issue**: Image too large\n\n```python\nfrom PIL import Image\n\ndef resize_if_needed(image_path, max_size=2048):\n    \"\"\"Resize image if dimensions exceed maximum.\"\"\"\n    img = Image.open(image_path)\n    \n    if max(img.size) > max_size:\n        img.thumbnail((max_size, max_size), Image.Resampling.LANCZOS)\n        resized_path = image_path.replace('.', '_resized.')\n        img.save(resized_path, quality=95)\n        return resized_path\n    \n    return image_path\n```\n\n**Issue**: Token limit exceeded\n\n```python\n# Reduce max_tokens or use low detail mode\nresponse = client.chat.completions.create(\n    model=\"gpt-5\",\n    messages=[...],\n    max_tokens=300  # Reduce from default\n)\n```\n\n**Issue**: Rate limit errors\n\n```python\nimport time\nfrom openai import RateLimitError\n\ndef analyze_with_retry(image_path, max_retries=3):\n    \"\"\"Analyze image with exponential backoff on rate limits.\"\"\"\n    for attempt in range(max_retries):\n        try:\n            return analyze_image(image_path)\n        except RateLimitError:\n            if attempt \u003C max_retries - 1:\n                wait_time = 2 ** attempt  # Exponential backoff\n                print(f\"Rate limit hit, waiting {wait_time}s...\")\n                time.sleep(wait_time)\n            else:\n                raise\n```\n\n\n## Best Practices\n\n### Prompt Engineering for Vision\n\n1. **Be specific**: \"Count the number of people wearing red shirts\" vs \"Analyze this image\"\n2. **Request structured output**: Ask for JSON, lists, or tables when appropriate\n3. **Provide context**: \"This is a medical diagram showing...\" helps the model understand\n4. **Use examples**: Show the format you want in your prompt\n\n### Image Quality Guidelines\n\n- Use clear, well-lit images\n- Ensure text is readable at original size\n- Avoid extreme angles or distortions\n- Crop to relevant content to save tokens\n- Use standard orientations (avoid rotated images)\n\n### Multi-Image Analysis\n\n- Order matters: Present images in logical sequence\n- Reference images explicitly: \"In the first image...\" \n- Limit to 10-20 images per request for best results\n- Use low detail for large batches of similar images\n\n\n## Quality Self-Check\n\nBefore returning results, verify:\n\n- [ ] Output is valid JSON (use `json.loads()` to validate)\n- [ ] All required fields are present and properly typed\n- [ ] API errors are caught and handled gracefully\n- [ ] Token usage is tracked and within limits\n- [ ] Image formats are supported (JPG, PNG, WEBP, GIF)\n- [ ] Base64 encoding is correct (no corruption)\n- [ ] Model name is valid and available\n- [ ] Results are consistent with the prompt request\n\n\n## Limitations\n\n### Vision Model Limitations\n\n- **Not for medical diagnosis**: Cannot interpret CT scans, X-rays, or provide medical advice\n- **Poor text recognition**: Struggles with rotated, upside-down, or very small text (\u003C 10pt)\n- **Non-Latin scripts**: Reduced accuracy for non-Latin alphabets and special characters\n- **Spatial reasoning**: Weak at precise localization tasks (e.g., chess positions, exact coordinates)\n- **Graph interpretation**: Cannot reliably distinguish line styles (solid vs dashed) or precise color gradients\n- **Object counting**: Provides approximate counts; may miss or double-count objects\n- **Panoramic\u002Ffisheye**: Distorted perspectives reduce accuracy\n- **Metadata loss**: Original EXIF data and exact dimensions are not preserved\n\n### Performance Considerations\n\n- **Latency**: High-detail mode takes longer to process\n- **Token costs**: Multiple images can quickly consume token budgets\n- **Rate limits**: Vision requests count toward TPM (tokens per minute) limits\n- **File size**: 20MB per image practical limit; 50MB total per request\n- **Batch size**: Over 20 images may degrade quality or hit timeouts\n\n","# OpenAI Vision Analysis Skill\n\n## Purpose\nThis skill enables image analysis, scene understanding, text extraction, and multi-frame comparison using OpenAI's vision-capable GPT models (e.g., `gpt-4o`, `gpt-5`). It supports single and multiple images analysis and sequential frames for temporal analysis.\n\n## When to Use\n- Analyzing image content (objects, scenes, colors, spatial relationships)\n- Extracting and reading text from images (OCR via vision models)\n- Comparing multiple images to detect differences or changes\n- Processing video frames to understand temporal progression\n- Generating detailed image descriptions or captions\n- Answering questions about visual content\n\n## Required Libraries\n\nThe following Python libraries are required:\n\n```python\nfrom openai import OpenAI\nimport base64\nimport json\nimport os\nfrom pathlib import Path\n```\n\n## Input Requirements\n- **File formats**: JPG, JPEG, PNG, WEBP, non-animated GIF\n- **Image quality**: Clear and legible; minimum 512×512px recommended\n- **File size**: Under 20MB per image recommended\n- **Maximum per request**: Up to 500 images, 50MB total payload\n- **URL or Base64**: Images can be provided as URLs or base64-encoded data\n\n## Output Schema\nAll analysis results should be returned as valid JSON conforming to this schema:\n\n```json\n{\n  \"success\": true,\n  \"model\": \"gpt-5\",\n  \"analysis\": \"Detailed description or analysis of the image content...\",\n  \"metadata\": {\n    \"image_count\": 1,\n    \"detail_level\": \"high\",\n    \"tokens_used\": 850,\n    \"processing_time_ms\": 1234\n  },\n  \"extracted_data\": {\n    \"objects\": [\"car\", \"person\", \"building\"],\n    \"text_found\": \"Sample text from image\",\n    \"colors\": [\"blue\", \"white\", \"gray\"],\n    \"scene_type\": \"urban street\"\n  },\n  \"warnings\": []\n}\n```\n\n### Field Descriptions\n\n- `success`: Boolean indicating whether the API call succeeded\n- `model`: The GPT model used for analysis (e.g., \"gpt-4o\", \"gpt-5\")\n- `analysis`: Complete textual analysis or description from the model\n- `metadata.image_count`: Number of images analyzed in this request\n- `metadata.detail_level`: Detail parameter used (\"low\", \"high\", or \"auto\")\n- `metadata.tokens_used`: Approximate token count for the request\n- `metadata.processing_time_ms`: Time taken to process the request\n- `extracted_data`: Structured information extracted from the image(s)\n- `warnings`: Array of issues or limitations encountered\n\n\n## Code Examples\n\n### Basic Image Analysis\n\n```python\nfrom openai import OpenAI\nimport base64\n\ndef analyze_image(image_path, prompt=\"What's in this image?\"):\n    \"\"\"Analyze a single image using GPT-5 Vision.\"\"\"\n    client = OpenAI(api_key=os.environ.get(\"OPENAI_API_KEY\"))\n    \n    # Read and encode image\n    with open(image_path, \"rb\") as image_file:\n        base64_image = base64.b64encode(image_file.read()).decode('utf-8')\n    \n    response = client.chat.completions.create(\n        model=\"gpt-5\",\n        messages=[\n            {\n                \"role\": \"user\",\n                \"content\": [\n                    {\"type\": \"text\", \"text\": prompt},\n                    {\n                        \"type\": \"image_url\",\n                        \"image_url\": {\n                            \"url\": f\"data:image\u002Fjpeg;base64,{base64_image}\"\n                        }\n                    }\n                ]\n            }\n        ],\n        max_tokens=300\n    )\n    \n    return response.choices[0].message.content\n```\n\n### Using Image URLs\n\n```python\nfrom openai import OpenAI\n\ndef analyze_image_url(image_url, prompt=\"Describe this image\"):\n    \"\"\"Analyze an image from a URL.\"\"\"\n    client = OpenAI(api_key=os.environ.get(\"OPENAI_API_KEY\"))\n    \n    response = client.chat.completions.create(\n        model=\"gpt-5\",\n        messages=[\n            {\n                \"role\": \"user\",\n                \"content\": [\n                    {\"type\": \"text\", \"text\": prompt},\n                    {\n                        \"type\": \"image_url\",\n                        \"image_url\": {\"url\": image_url}\n                    }\n                ]\n            }\n        ]\n    )\n    \n    return response.choices[0].message.content\n```\n\n### Multiple Images Analysis\n\n```python\nfrom openai import OpenAI\nimport base64\n\ndef analyze_multiple_images(image_paths, prompt=\"Compare these images\"):\n    \"\"\"Analyze multiple images in a single request.\"\"\"\n    client = OpenAI(api_key=os.environ.get(\"OPENAI_API_KEY\"))\n    \n    # Build content array with text and all images\n    content = [{\"type\": \"text\", \"text\": prompt}]\n    \n    for image_path in image_paths:\n        with open(image_path, \"rb\") as image_file:\n            base64_image = base64.b64encode(image_file.read()).decode('utf-8')\n        \n        content.append({\n            \"type\": \"image_url\",\n            \"image_url\": {\n                \"url\": f\"data:image\u002Fjpeg;base64,{base64_image}\"\n            }\n        })\n    \n    response = client.chat.completions.create(\n        model=\"gpt-5\",\n        messages=[{\"role\": \"user\", \"content\": content}],\n        max_tokens=500\n    )\n    \n    return response.choices[0].message.content\n```\n\n### Full Analysis with JSON Output\n\n```python\nfrom openai import OpenAI\nimport base64\nimport json\nimport time\n\ndef analyze_image_to_json(image_path, prompt=\"Analyze this image\"):\n    \"\"\"Analyze image and return structured JSON output.\"\"\"\n    client = OpenAI(api_key=os.environ.get(\"OPENAI_API_KEY\"))\n    \n    start_time = time.time()\n    warnings = []\n    \n    try:\n        # Read and encode image\n        with open(image_path, \"rb\") as image_file:\n            base64_image = base64.b64encode(image_file.read()).decode('utf-8')\n        \n        # Make API call\n        response = client.chat.completions.create(\n            model=\"gpt-5\",\n            messages=[\n                {\n                    \"role\": \"user\",\n                    \"content\": [\n                        {\"type\": \"text\", \"text\": prompt},\n                        {\n                            \"type\": \"image_url\",\n                            \"image_url\": {\n                                \"url\": f\"data:image\u002Fjpeg;base64,{base64_image}\",\n                                \"detail\": \"high\"\n                            }\n                        }\n                    ]\n                }\n            ],\n            max_tokens=500\n        )\n        \n        analysis = response.choices[0].message.content\n        tokens_used = response.usage.total_tokens\n        processing_time = int((time.time() - start_time) * 1000)\n        \n        result = {\n            \"success\": True,\n            \"model\": \"gpt-5\",\n            \"analysis\": analysis,\n            \"metadata\": {\n                \"image_count\": 1,\n                \"detail_level\": \"high\",\n                \"tokens_used\": tokens_used,\n                \"processing_time_ms\": processing_time\n            },\n            \"extracted_data\": {},\n            \"warnings\": warnings\n        }\n        \n    except Exception as e:\n        result = {\n            \"success\": False,\n            \"model\": \"gpt-5\",\n            \"analysis\": \"\",\n            \"metadata\": {\n                \"image_count\": 0,\n                \"detail_level\": \"high\",\n                \"tokens_used\": 0,\n                \"processing_time_ms\": 0\n            },\n            \"extracted_data\": {},\n            \"warnings\": [f\"API call failed: {str(e)}\"]\n        }\n    \n    return result\n\n# Usage\nresult = analyze_image_to_json(\"photo.jpg\", \"Describe what you see in detail\")\nprint(json.dumps(result, indent=2))\n```\n\n### Batch Processing with Sequential Frames\n\n```python\nfrom openai import OpenAI\nimport base64\nfrom pathlib import Path\n\ndef process_video_frames(frames_directory, analysis_prompt):\n    \"\"\"Process sequential video frames for temporal analysis.\"\"\"\n    client = OpenAI(api_key=os.environ.get(\"OPENAI_API_KEY\"))\n    \n    image_extensions = {'.jpg', '.jpeg', '.png', '.webp'}\n    frame_paths = sorted([\n        f for f in Path(frames_directory).iterdir() \n        if f.suffix.lower() in image_extensions\n    ])\n    \n    # Analyze frames in groups (e.g., 5 frames at a time)\n    batch_size = 5\n    results = []\n    \n    for i in range(0, len(frame_paths), batch_size):\n        batch = frame_paths[i:i+batch_size]\n        \n        # Build content with all frames in batch\n        content = [{\"type\": \"text\", \"text\": analysis_prompt}]\n        \n        for frame_path in batch:\n            with open(frame_path, \"rb\") as f:\n                base64_image = base64.b64encode(f.read()).decode('utf-8')\n            \n            content.append({\n                \"type\": \"image_url\",\n                \"image_url\": {\n                    \"url\": f\"data:image\u002Fjpeg;base64,{base64_image}\",\n                    \"detail\": \"low\"  # Use low detail for video frames to save tokens\n                }\n            })\n        \n        response = client.chat.completions.create(\n            model=\"gpt-5\",\n            messages=[{\"role\": \"user\", \"content\": content}],\n            max_tokens=800\n        )\n        \n        results.append({\n            \"batch_index\": i \u002F\u002F batch_size,\n            \"frame_range\": f\"{batch[0].name} to {batch[-1].name}\",\n            \"analysis\": response.choices[0].message.content\n        })\n    \n    return results\n```\n\n### Text Extraction from Images (OCR Alternative)\n\n```python\nfrom openai import OpenAI\nimport base64\n\ndef extract_text_with_gpt(image_path):\n    \"\"\"Extract text from image using GPT Vision as OCR alternative.\"\"\"\n    client = OpenAI(api_key=os.environ.get(\"OPENAI_API_KEY\"))\n    \n    with open(image_path, \"rb\") as image_file:\n        base64_image = base64.b64encode(image_file.read()).decode('utf-8')\n    \n    response = client.chat.completions.create(\n        model=\"gpt-5\",\n        messages=[\n            {\n                \"role\": \"user\",\n                \"content\": [\n                    {\n                        \"type\": \"text\", \n                        \"text\": \"Extract all text from this image. Return only the text content, preserving the layout and structure.\"\n                    },\n                    {\n                        \"type\": \"image_url\",\n                        \"image_url\": {\n                            \"url\": f\"data:image\u002Fjpeg;base64,{base64_image}\",\n                            \"detail\": \"high\"\n                        }\n                    }\n                ]\n            }\n        ],\n        max_tokens=1000\n    )\n    \n    return response.choices[0].message.content\n```\n\n\n## Model Selection and Configuration\n\n### Available Models\n\n```python\n# GPT-4o - Best for general vision tasks, fast and cost-effective\nmodel = \"gpt-4o\"\n\n# GPT-5-nano - Faster and cheaper for simple vision tasks\nmodel = \"gpt-5-nano\"\n\n# GPT-5 - More capable for complex reasoning\nmodel = \"gpt-5\"\n```\n\n### Detail Level Configuration\n\nControl how much visual detail the model processes:\n\n```python\n# Low detail - 512×512px resolution, fewer tokens, faster\n\"image_url\": {\n    \"url\": image_url,\n    \"detail\": \"low\"\n}\n\n# High detail - Full resolution with tiling, more tokens, better accuracy\n\"image_url\": {\n    \"url\": image_url,\n    \"detail\": \"high\"\n}\n\n# Auto - Model chooses appropriate detail level\n\"image_url\": {\n    \"url\": image_url,\n    \"detail\": \"auto\"\n}\n```\n\n**When to use each detail level:**\n- **Low**: Video frames, simple scene classification, color\u002Fshape detection\n- **High**: Text extraction, detailed object detection, fine-grained analysis\n- **Auto**: General purpose when unsure; model optimizes cost vs. quality\n\n\n## Token Cost Management\n\n### Understanding Image Tokens\n\nImage tokens count toward your request limits and costs:\n\n- **Low detail**: Fixed ~85 tokens per image (gpt-5)\n- **High detail**: Base tokens + tile tokens based on image dimensions\n  - Images are scaled to fit within 2048×2048px\n  - Divided into 512×512px tiles\n  - Each tile costs additional tokens\n\n### Cost Calculation Examples\n\n```python\n# For gpt-5 with high detail:\n# - Base: 85 tokens\n# - Per tile: 170 tokens\n# - Example: 1024×1024 image = 85 + (2×2 tiles × 170) = 765 tokens\n\ndef estimate_tokens_high_detail(width, height):\n    \"\"\"Estimate token cost for high-detail image (gpt-5).\"\"\"\n    # Scale to fit within 2048×2048\n    scale = min(2048 \u002F width, 2048 \u002F height, 1.0)\n    scaled_w = int(width * scale)\n    scaled_h = int(height * scale)\n    \n    # Calculate tiles (512×512)\n    tiles_x = (scaled_w + 511) \u002F\u002F 512\n    tiles_y = (scaled_h + 511) \u002F\u002F 512\n    total_tiles = tiles_x * tiles_y\n    \n    # Token calculation\n    base_tokens = 85\n    tile_tokens = total_tiles * 170\n    \n    return base_tokens + tile_tokens\n```\n\n### Cost Optimization Strategies\n\n1. **Use low detail for video frames** - Temporal analysis doesn't need high resolution\n2. **Resize large images before uploading** - Reduce dimensions to 1024×1024 if high detail not needed\n3. **Batch related questions** - Analyze multiple aspects in one API call\n4. **Cache analysis results** - Store results for repeated processing\n\n\n## Advanced Use Cases\n\n### Image Comparison\n\n```python\ndef compare_images(image1_path, image2_path):\n    \"\"\"Compare two images and identify differences.\"\"\"\n    client = OpenAI(api_key=os.environ.get(\"OPENAI_API_KEY\"))\n    \n    images = []\n    for path in [image1_path, image2_path]:\n        with open(path, \"rb\") as f:\n            base64_image = base64.b64encode(f.read()).decode('utf-8')\n            images.append(base64_image)\n    \n    response = client.chat.completions.create(\n        model=\"gpt-5\",\n        messages=[\n            {\n                \"role\": \"user\",\n                \"content\": [\n                    {\n                        \"type\": \"text\",\n                        \"text\": \"Compare these two images. List all differences you observe, including changes in objects, colors, positions, or any other visual elements.\"\n                    },\n                    {\"type\": \"image_url\", \"image_url\": {\"url\": f\"data:image\u002Fjpeg;base64,{images[0]}\"}},\n                    {\"type\": \"image_url\", \"image_url\": {\"url\": f\"data:image\u002Fjpeg;base64,{images[1]}\"}}\n                ]\n            }\n        ]\n    )\n    \n    return response.choices[0].message.content\n```\n\n### Structured Data Extraction\n\n```python\ndef extract_structured_data(image_path, schema_description):\n    \"\"\"Extract structured information from image based on schema.\"\"\"\n    client = OpenAI(api_key=os.environ.get(\"OPENAI_API_KEY\"))\n    \n    with open(image_path, \"rb\") as f:\n        base64_image = base64.b64encode(f.read()).decode('utf-8')\n    \n    prompt = f\"\"\"Analyze this image and extract information in JSON format following this schema:\n{schema_description}\n\nReturn only valid JSON, no additional text.\"\"\"\n    \n    response = client.chat.completions.create(\n        model=\"gpt-5\",\n        messages=[\n            {\n                \"role\": \"user\",\n                \"content\": [\n                    {\"type\": \"text\", \"text\": prompt},\n                    {\n                        \"type\": \"image_url\",\n                        \"image_url\": {\n                            \"url\": f\"data:image\u002Fjpeg;base64,{base64_image}\",\n                            \"detail\": \"high\"\n                        }\n                    }\n                ]\n            }\n        ],\n        response_format={\"type\": \"json_object\"}\n    )\n    \n    return json.loads(response.choices[0].message.content)\n\n# Usage example\nschema = \"\"\"\n{\n  \"products\": [{\"name\": string, \"price\": number, \"quantity\": number}],\n  \"total\": number,\n  \"date\": string\n}\n\"\"\"\ndata = extract_structured_data(\"receipt.jpg\", schema)\n```\n\n\n## Error Handling\n\n### Common Issues and Solutions\n\n**Issue**: API authentication failed\n\n```python\n# Verify API key is set\nimport os\napi_key = os.environ.get(\"OPENAI_API_KEY\")\nif not api_key:\n    raise ValueError(\"OPENAI_API_KEY environment variable not set\")\n```\n\n**Issue**: Image too large\n\n```python\nfrom PIL import Image\n\ndef resize_if_needed(image_path, max_size=2048):\n    \"\"\"Resize image if dimensions exceed maximum.\"\"\"\n    img = Image.open(image_path)\n    \n    if max(img.size) > max_size:\n        img.thumbnail((max_size, max_size), Image.Resampling.LANCZOS)\n        resized_path = image_path.replace('.', '_resized.')\n        img.save(resized_path, quality=95)\n        return resized_path\n    \n    return image_path\n```\n\n**Issue**: Token limit exceeded\n\n```python\n# Reduce max_tokens or use low detail mode\nresponse = client.chat.completions.create(\n    model=\"gpt-5\",\n    messages=[...],\n    max_tokens=300  # Reduce from default\n)\n```\n\n**Issue**: Rate limit errors\n\n```python\nimport time\nfrom openai import RateLimitError\n\ndef analyze_with_retry(image_path, max_retries=3):\n    \"\"\"Analyze image with exponential backoff on rate limits.\"\"\"\n    for attempt in range(max_retries):\n        try:\n            return analyze_image(image_path)\n        except RateLimitError:\n            if attempt \u003C max_retries - 1:\n                wait_time = 2 ** attempt  # Exponential backoff\n                print(f\"Rate limit hit, waiting {wait_time}s...\")\n                time.sleep(wait_time)\n            else:\n                raise\n```\n\n\n## Best Practices\n\n### Prompt Engineering for Vision\n\n1. **Be specific**: \"Count the number of people wearing red shirts\" vs \"Analyze this image\"\n2. **Request structured output**: Ask for JSON, lists, or tables when appropriate\n3. **Provide context**: \"This is a medical diagram showing...\" helps the model understand\n4. **Use examples**: Show the format you want in your prompt\n\n### Image Quality Guidelines\n\n- Use clear, well-lit images\n- Ensure text is readable at original size\n- Avoid extreme angles or distortions\n- Crop to relevant content to save tokens\n- Use standard orientations (avoid rotated images)\n\n### Multi-Image Analysis\n\n- Order matters: Present images in logical sequence\n- Reference images explicitly: \"In the first image...\" \n- Limit to 10-20 images per request for best results\n- Use low detail for large batches of similar images\n\n\n## Quality Self-Check\n\nBefore returning results, verify:\n\n- [ ] Output is valid JSON (use `json.loads()` to validate)\n- [ ] All required fields are present and properly typed\n- [ ] API errors are caught and handled gracefully\n- [ ] Token usage is tracked and within limits\n- [ ] Image formats are supported (JPG, PNG, WEBP, GIF)\n- [ ] Base64 encoding is correct (no corruption)\n- [ ] Model name is valid and available\n- [ ] Results are consistent with the prompt request\n\n\n## Limitations\n\n### Vision Model Limitations\n\n- **Not for medical diagnosis**: Cannot interpret CT scans, X-rays, or provide medical advice\n- **Poor text recognition**: Struggles with rotated, upside-down, or very small text (\u003C 10pt)\n- **Non-Latin scripts**: Reduced accuracy for non-Latin alphabets and special characters\n- **Spatial reasoning**: Weak at precise localization tasks (e.g., chess positions, exact coordinates)\n- **Graph interpretation**: Cannot reliably distinguish line styles (solid vs dashed) or precise color gradients\n- **Object counting**: Provides approximate counts; may miss or double-count objects\n- **Panoramic\u002Ffisheye**: Distorted perspectives reduce accuracy\n- **Metadata loss**: Original EXIF data and exact dimensions are not preserved\n\n### Performance Considerations\n\n- **Latency**: High-detail mode takes longer to process\n- **Token costs**: Multiple images can quickly consume token budgets\n- **Rate limits**: Vision requests count toward TPM (tokens per minute) limits\n- **File size**: 20MB per image practical limit; 50MB total per request\n- **Batch size**: Over 20 images may degrade quality or hit timeouts\n\n",{"name":5,"description":7},"main","7eeb258ada72b7f58119b0e7a2c00ea89d9beedd","tasks-no-skills\u002Fpedestrian-traffic-counting\u002Fenvironment\u002Fskills\u002Fgpt-multimodal","https:\u002F\u002Fgithub.com\u002FWaaaalnut\u002Fskillsbench","https:\u002F\u002Fgithub.com\u002FWaaaalnut","student of Zhejiang University","2026-04-21 23:16:23",{"source":8,"skillName":5,"commands":41,"prompt":51,"zipAvailable":52},[42,45,48],{"tool":43,"command":44},"npx","npx skills add Waaaalnut\u002Fskillsbench --skill gpt-multimodal",{"tool":46,"command":47},"bunx","bunx skills add Waaaalnut\u002Fskillsbench --skill gpt-multimodal",{"tool":49,"command":50},"pnpm","pnpm dlx skills add Waaaalnut\u002Fskillsbench --skill gpt-multimodal","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FWaaaalnut\u002Fskillsbench\u002Fblob\u002Fmain\u002Ftasks-no-skills\u002Fpedestrian-traffic-counting\u002Fenvironment\u002Fskills\u002Fgpt-multimodal\u002FSKILL.md",true,[],[55,69,84,112,134,153],{"skillNo":56,"name":57,"nameZh":6,"description":6,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":58,"categories":59,"keywords":65,"publishedAt":28},"01M6XEWMJ6VY3ZC53MPDCPG3GH","data-reconciliation","https:\u002F\u002Fgithub.com\u002FWaaaalnut\u002Fskillsbench\u002Fblob\u002Fmain\u002Ftasks\u002Fxlsx-recover-data\u002Fenvironment\u002Fskills\u002Fdata-reconciliation\u002FSKILL.md",[60],{"categoryNo":61,"categoryName":62,"taxonomyType":63,"taxonomyTypeDesc":64},"01JBSEEDCAT000000000000033","数据分析师","PROFESSION","适用职业",[66,67,68],"data reconciliation","reconciliation","data",{"skillNo":70,"name":71,"nameZh":6,"description":72,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":73,"categories":74,"keywords":75,"publishedAt":28},"01M6XEW8MNPMF1KRXV6ATGFRQF","speech-to-text","Transcribe video to timestamped text using Whisper tiny model (pre-installed).","https:\u002F\u002Fgithub.com\u002FWaaaalnut\u002Fskillsbench\u002Fblob\u002Fmain\u002Ftasks\u002Fvideo-tutorial-indexer\u002Fenvironment\u002Fskills\u002Fspeech-to-text\u002FSKILL.md",[],[76,77,78,79,80,81,82,83],"speech","text","speech to text","transcribe video","whisper","model","Video","transcribe",{"skillNo":85,"name":86,"nameZh":6,"description":87,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":88,"categories":89,"keywords":95,"publishedAt":28},"01M6XEW3FQZP0T1RKQQJFXVY9T","video-processor","Process videos by removing segments and concatenating remaining parts. Use when you need to remove detected pauses\u002Fopenings from videos, create highlight reels, or batch process segment removals using ffmpeg filter_complex.","https:\u002F\u002Fgithub.com\u002FWaaaalnut\u002Fskillsbench\u002Fblob\u002Fmain\u002Ftasks\u002Fvideo-silence-remover\u002Fenvironment\u002Fskills\u002Fvideo-processor\u002FSKILL.md",[90],{"categoryNo":91,"categoryName":92,"taxonomyType":93,"taxonomyTypeDesc":94},"01JBSEEDCAT000000000000016","图像与多媒体","SCENARIO","使用场景",[82,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111],"processor","video processor","batch","create","reels","when","parts","segments","process","videos","you","segment","filter","ffmpeg","use","remove",{"skillNo":113,"name":114,"nameZh":6,"description":115,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":116,"categories":117,"keywords":119,"publishedAt":28},"01M6XEVXJ5KZ0NTY3XZKRY4AS1","silence-detector","Detect initial silence segments in audio\u002Fvideo using energy-based analysis. Use when you need to find low-energy periods at the start of recordings (title slides, setup time, pre-roll silence).","https:\u002F\u002Fgithub.com\u002FWaaaalnut\u002Fskillsbench\u002Fblob\u002Fmain\u002Ftasks\u002Fvideo-silence-remover\u002Fenvironment\u002Fskills\u002Fsilence-detector\u002FSKILL.md",[118],{"categoryNo":91,"categoryName":92,"taxonomyType":93,"taxonomyTypeDesc":94},[120,121,122,123,101,106,124,125,126,127,128,103,129,130,82,131,132,133,110],"detector","silence detector","title","slides","time","energy","at","recordings","detect","find","Audio","analysis","start","setup",{"skillNo":135,"name":136,"nameZh":6,"description":137,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":138,"categories":139,"keywords":140,"publishedAt":28},"01M6XEVT3NGRN1FT7JPMAPXS6M","segment-combiner","Combine multiple segment detection results into a unified list. Use when you need to merge segments from different detectors, prepare removal lists for video processing, or consolidate detection outputs.","https:\u002F\u002Fgithub.com\u002FWaaaalnut\u002Fskillsbench\u002Fblob\u002Fmain\u002Ftasks\u002Fvideo-silence-remover\u002Fenvironment\u002Fskills\u002Fsegment-combiner\u002FSKILL.md",[],[107,141,82,142,143,144,145,146,101,147,148,103,106,149,150,110,151,152],"segment combiner","detection","prepare","merge","video processing","for","outputs","combine","processing","consolidate","results","list",{"skillNo":154,"name":155,"nameZh":6,"description":156,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":9,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":157,"categories":158,"keywords":162,"publishedAt":28},"01M6XEVMPVASGS2KX86QK4FMT7","report-generator","Generate compression reports for video processing. Use when you need to create structured JSON reports with duration statistics, compression ratios, and segment details after video processing.","https:\u002F\u002Fgithub.com\u002FWaaaalnut\u002Fskillsbench\u002Fblob\u002Fmain\u002Ftasks\u002Fvideo-silence-remover\u002Fenvironment\u002Fskills\u002Freport-generator\u002FSKILL.md",[159],{"categoryNo":160,"categoryName":161,"taxonomyType":63,"taxonomyTypeDesc":64},"01JBSEEDCAT000000000000034","行政与办公人员",[163,164,165,107,101,146,106,166,149,167,168,169,110,145,170,82,171,172,99],"generator","report generator","report","with","statistics","structured","reports","json","compression","generate"]