Vision Cortex (Perception Gateway)

Analyze uploaded images via NATS RPC to identify objects and extract metadata.

3|2|Updated Jan 24, 2026

Search Tags:#data extraction #multimodal #ai agent #vision #perception #image analysis

Ready to run

npx skills add https://github.com/zaebee/aura --skill vision-cortex-perception-gateway

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: Vision Cortex (Perception Gateway)
Source: https://github.com/zaebee/aura/tree/main/api-gateway
Command: npx skills add https://github.com/zaebee/aura --skill vision-cortex-perception-gateway

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This Skill processes visual data, translating raw images into structured information and identifying key assets, enabling AI agents to understand and interact with the visual world.

Core Features & Use Cases

Image Analysis: Analyzes uploaded images to identify objects, categorize them, and extract relevant metadata.
Multimodal Input: Accepts image files via multipart/form-data requests.
Use Case: An agent needs to identify a specific car model from a photo for a negotiation. It sends the image to the Vision Cortex, which returns the car's make, model, and estimated value.

Quick Start

Use the Vision Cortex skill to analyze the attached image file 'car.jpg' with a focus on identifying the vehicle.

Frequently Asked Questions about Vision Cortex (Perception Gateway)

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I extract structured data and metadata from an image upload?▼

Vision Cortex performs multimodal image analysis by processing uploaded image files through a NATS RPC worker swarm. It identifies objects, extracts metadata, and returns structured insights for AI agents to understand visual data.

How do I analyze an uploaded image file to identify objects for an AI agent?▼

Submit image files via multipart/form-data requests with optional focus hints to analyze visual data. The worker swarm identifies objects and extracts metadata, returning structured insights directly usable by AI agents.

Can I use visual data analysis to identify specific assets like a car model from a photo?▼

Yes, you can analyze a photo to identify specific assets like a car model. Send the image with a focus hint, and the perception gateway returns structured insights including the vehicle's make, model, and estimated value for agent negotiation.

Does the image analysis worker enforce request size limits and identity verification?▼

Yes, the perception gateway enforces safety protocols including request size limits and identity verification. These measures ensure secure and controlled processing of image uploads through the NATS RPC worker swarm.

What's the best way to process visual data for AI agents using a worker swarm?▼

Route multipart/form-data image uploads via NATS RPC to a worker swarm to process visual data for AI agents. This approach extracts metadata and returns structured insights while enforcing request size and identity verification protocols.