Image Analysis Skills

Skills for analyzing, transforming, and previewing images, including sandbox-generated artifacts.

/api/skills/image

1 Skill

image_analysis

Analyze, transform, and preview images using analyze_image and artifact-aware tools.

/api/skills/image/image_analysis
Skill python
Content Type
Static (Markdown)
Language
python
Widgets
n/a

Trigger Keywords

imageimagesanalyzecropgridartifactpreviewsandboxdocument

Instructions Preview

# Image Analysis Skill Use this skill when you need to: - Analyze or transform images and PDFs with ImageMagick via the **`analyze_image`** tool. - Work with images produced inside the Daytona sandbox (for example, plots saved to `/home/daytona/out`). - Bring sandbox-generated images back into the **model context** so they are visible as images, not just file paths. - Surface Vertesia documents or objects (e.g. PDFs, images) as images the model can see. This skill explains how to combine: - `execute_shell` + `write_artifact` + `/home/daytona/out` for code-generated images. - `analyze_image` for ImageMagick operations and grid overlays. - `view_image` to surface either: - artifact images under `out/*`, or - images/visual documents from Vertesia objects (via `object_id`), as conversation-level images that the model can see. --- ## 1. Analyzing existing documents and images When you have an existing Vertesia object (document or image), you have two options: 1. Use the `objec...