analyze_image

Instructions

Implementation Reference

• src/yaparai/tools/ai.py:39-62 (handler)

  The core handler function for the 'analyze_image' tool. It accepts `image_url` (str) and `prompt` (str, default 'Describe this image in detail'), creates a YaparAIClient, and calls `gemini_analyze_image` to send the request to the Gemini Vision AI API.

  async def analyze_image(
      image_url: str,
      prompt: str = "Describe this image in detail",
  ) -> dict:
      """
      Analyze an image using Gemini Vision AI.
  
      Upload an image and ask questions about it. Can identify objects,
      read text, describe scenes, analyze compositions, and more.
      Cost: ~2 credits.
  
      Args:
          image_url: URL of the image to analyze
          prompt: Question or instruction about the image
              (e.g., "What product is shown?", "Read the text in this image")
  
      Returns:
          Dict with analysis text and details.
      """
      client = YaparAIClient()
      return await client.gemini_analyze_image({
          "image_url": image_url,
          "prompt": prompt,
      })

• src/yaparai/tools/ai.py:39-62 (schema)

  The function signature and docstring define the input schema: `image_url: str` (required) and `prompt: str` (optional, defaulting to 'Describe this image in detail'). The return is a dict with analysis text and details.

  async def analyze_image(
      image_url: str,
      prompt: str = "Describe this image in detail",
  ) -> dict:
      """
      Analyze an image using Gemini Vision AI.
  
      Upload an image and ask questions about it. Can identify objects,
      read text, describe scenes, analyze compositions, and more.
      Cost: ~2 credits.
  
      Args:
          image_url: URL of the image to analyze
          prompt: Question or instruction about the image
              (e.g., "What product is shown?", "Read the text in this image")
  
      Returns:
          Dict with analysis text and details.
      """
      client = YaparAIClient()
      return await client.gemini_analyze_image({
          "image_url": image_url,
          "prompt": prompt,
      })

• src/yaparai/server.py:145-147 (registration)

  Registration of the tool via `mcp.tool(analyze_image)` in the FastMCP server under the '# AI Tools (2)' section.

  # AI Tools (2)
  mcp.tool(generate_text)
  mcp.tool(analyze_image)

• src/yaparai/client.py:185-187 (helper)

  The API client method `gemini_analyze_image` that sends the POST request to '/v1/ai/gemini/analyze-image' with the payload containing image_url and prompt.

  async def gemini_analyze_image(self, payload: dict) -> dict:
      """Gemini Vision image analysis."""
      return await self._request("POST", "/v1/ai/gemini/analyze-image", json=payload)

• src/yaparai/server.py:55-58 (helper)

  Import of the `analyze_image` function from `yaparai.tools.ai` into the main server module.

  from yaparai.tools.ai import (
      generate_text,
      analyze_image,
  )