Releases: askui/vision-agent
Releases · askui/vision-agent
v0.1.11
What's Changed
- feat: add hub tool by @adi-wan-askui in #12
New Contributors
- @adi-wan-askui made their first contribution in #12
Full Changelog: v0.1.9...v0.1.11
Reporting
- Feature to generate HTML report of an automation run
Integrate ShowUI HF Model
showlab/ShowUI-2B
now available as click model via HF Spaces
Hugging Face Integration
- Integrate Hugging Face Spaces models for click command
Bugfixes
- add missing requests dependency
- fix agent.keyboard and add type hints for keyboard commands
AskUI Integration
- Integrates with AskUI as a model provider
- Adds AskUI PTA and OCR models
- Adds model routing to execute click commands with different models from different model providers
- Checks token availability for model providers and raises AutomationError on missing auth
Direct Tool Use Docs and Clipboard
- integrated pyperclip for clipboard access
- agent os controlls now behind tools interface
agent.tools.os...
- improved docs on tools
Keyboard Fixes
- Map keysym key names to agent os names -> fixes issues with keyboard use in act()
Toolbox and Bugfixes
- Move tools like webbrowser under a tools attribute (breaking change)
- improved docs
- disable default debugDraw flag
Improved Logging and Display Selection Fixes
- Improved logging for Agent activity
- Set default log_level to INFO
- Improved Readme for installation and added Wayland warning
- Fixes for display selection