aura vision

Analyze an image or PDF via a vision-capable LLM.

Syntax

aura vision <file> [instruction]

Description

Sends an image or PDF to a vision-capable LLM for analysis. If no instruction is given, the model extracts text or describes what it sees.

Requires a vision agent configured in features/vision.yaml.

Examples

# Describe an image
aura vision screenshot.png

# Extract text from a PDF
aura vision document.pdf "Extract all text from this document"

# Analyze a diagram
aura vision diagram.jpg "Describe the architecture shown here"

Plus all global flags.