AgentSkillsCN

image-describer

使用 Gemini 3 Flash 描述或分析图像内容。当您需要理解图像或 PDF 文件中的信息（例如图标、截图、图表），以便推进相关任务时，可使用此技能。

SKILL.md
--- frontmatter
name: image-describer
description: Describe or analyze image content using Gemini 3 Flash. Use this skill when you need to understand what is in an image or pdf file (e.g., icons, screenshots, diagrams) to proceed with a task.

Image Describer

This skill allows you to analyze image files by providing a prompt and the file path. It uses the gemini-3-flash-preview model for fast and accurate vision analysis.

Workflow for Agents

When you encounter an image file and need to understand its content:

  1. •Identify the absolute path of the image file.
  2. •Decide what information you need from the image (e.g., general description, text extraction, UI component identification).
  3. •Execute the describe.py script.

Example Usage

bash
python3 scripts/describe.py --prompt "Extract all text from this image." --image "screenshot.png"

Specific Use Cases

  • •OCR: python3 scripts/describe.py --prompt "Extract all text from this image." --image "screenshot.png"
  • •UI Analysis: python3 scripts/describe.py --prompt "Identify the main UI components and their layout." --image "mockup.jpg"
  • •Icon Identification: python3 scripts/describe.py --prompt "What does this icon represent?" --image "icon.svg"

Resource Details

  • •scripts/describe.py: Vertex AI SDK (google-genai) を使用して、画像や PDF を解析する Python スクリプト。