skill

1# vision-skills
2 
3Ten native DSH tools give a text-only agent eyes. Use these structured tools
4directly; do not shell out to the bundled Python scripts or reproduce their
5implementation. Vision API credentials and model settings are managed by the
6plugin, so tool calls do not receive credentials.
7 
8The visual execution schemas are mounted only for the current Agent after this
9Skill is loaded. A normal `skill` call activates them for the next model step.
10If this content arrived through a direct `/vision-skills` invocation and the
11visual tools are still absent, call `vision_toolkit_activate` once. Do not call
12that bootstrap when the visual tools are already present.
13 
14Pick the tool by the question you are answering:
15 
16| Question | Tool |
17|---|---|
18| "What does this image show / say?" | `vision_glance` |
19| "Where is X?" — a thing you can name | `vision_ground` |
20| "Where are all the Xs?" — every instance of a kind | `vision_detect` |
21| "What is its exact shape, size, offset?" | `vision_trace` |
22| "Cut this box out as its own image file" | `vision_crop` |
23| "OCR this long screenshot / scrolling page / chat history" | `vision_long_screenshot_ocr` |
24| "Extract the icon/logo foreground as transparent PNG — manual region or auto (cropped+scaled screenshots)" | `vision_extract_foreground` |
25| "Turn this HTML file into a screenshot" | `vision_html_screenshot` |
26| "Which colours dominate a region, and which palette value fits it?" | `vision_dominant_colors` |
27…