ASI

Want your AI Agent to understand images? Build an Image Analysis Agent in 7 steps 👇

1. Create a hosted Agent on http://Agentverse.ai

Start with a new hosted Agent and add:
'http://agent.py
image_analysis.py'

2. Add the Agent Chat Protocol

Handle:

ChatMessage
ChatAcknowledgement
TextContent
ResourceContent

This gives ASI:One a standard way to send the prompt + image attachment to your Agent.

3. Enable attachments

When a session starts, advertise:
{"attachments": "true"}

Now ASI:One Chat knows your Agent accepts files.

4. Extract the image URL

Read the HTTPS URL from ResourceContent and pass it into your image-analysis function.

5. Connect a multimodal model

Send:
text prompt + image URL → multimodal model → analysis

6. Publish the protocol + describe the capability

Use:
agent.include(chat_proto, publish_manifest=True)

Then add a clear README explaining exactly what your Agent can analyse so it can be discovered correctly.

7. Test it through http://asi1.ai 

Start the Agent → open its profile → select Chat with Agent → attach an image → ask: 'Please analyse this image'

Your flow is now:
ASI:One → image attachment → Agent Chat Protocol → multimodal model → analysis → ASI:One

Build it step-by-step → https://innovationlab.fetch.ai/resources/docs/examples/chat-protocol/image-analysis-agent