Menu
Vision capabilities: screen analysis, OCR, image description, UI element finding, and more.
The review-first prompt is selected by default. You can switch to a direct command or download a local copy.
Read SKILL.md and any companion files shown by SkillsMP before deciding whether to install.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
A direct command skips the review prompt. Inspect the source before running it.
npx skills add https://github.com/SanthoshReddy352/Friday_Linux --skill visionThe command stays on one line. Scroll horizontally to inspect it before copying.
Prefer a local copy? Download the files currently available to SkillsMP.
Sandboxed Python and Bash code execution for calculations, scripts, and data tasks.
MCP (Model Context Protocol) client — connects to local or remote MCP servers and registers their tools dynamically.
Show, save, forget, wipe, and export FRIDAY's memory about you
Based on SOC occupation classification
Showing SKILL.md
| name | vision |
| description | Vision capabilities: screen analysis, OCR, image description, UI element finding, and more. |
| plugin_module | modules/vision |
| capabilities | [{"name":"analyze_screen","description":"Take a screenshot and describe what's on the screen.","aliases":["what's on my screen","analyze my screen","what do you see"]},{"name":"describe_image","description":"Describe an image by file path or public URL.","aliases":["describe this image","what is in this image","analyze this photo","describe the image at"]},{"name":"read_text_from_image","description":"OCR — extract text from a screenshot or image.","aliases":["read the screen","extract text from screen","what does this say"]},{"name":"summarize_screen","description":"Give a high-level summary of what's currently on screen.","aliases":["summarize my screen","what am I looking at"]},{"name":"analyze_clipboard_image","description":"Analyze an image copied to the clipboard.","aliases":["analyze clipboard","explain this image"]},{"name":"find_ui_element","description":"Find a UI element on screen by description (button, menu, input field).","aliases":["where is the button","find the settings menu","locate the button"]}] |
Uses a local vision-language model (SmolVLM2 / LLaVA) to understand images and your screen.
# config.yaml — vision section
vision:
model_path: "models/smolvlm2.gguf"
mmproj_path: "models/smolvlm2-mmproj.gguf"
Or install llava-1.6 via llama.cpp for the fallback.
Friday, what's on my screen?
Friday, read the text on my screen
Friday, describe the image at /home/user/diagram.png
Friday, analyze the clipboard image
Friday, where is the save button?
slow (VLM inference takes 4–15s on CPU).describe_image downloads public URLs to a temp file before processing.analyze_screen requires screenshot capability (see modules/system_control/SKILL.md).