Release v0.7.3

Release v0.7.3

Highlights

Critical bug fix for browser automation. The image_result_formatter plugin was causing 500 Internal Server Error when processing screenshots. This release fixes the message sequence to comply with LLM API requirements.

What's Changed

🐛 Bug Fixes

• Fixed 500 Internal Server Error in image_result_formatter plugin
  • Changed image message role from 'assistant' to 'user'
  • LLM APIs (OpenAI, Anthropic, Gemini) require alternating user/assistant messages
  • Previous implementation created two consecutive assistant messages, causing API rejection

✨ Improvements

• Synced browser agent with browser-agent project
  • take_screenshot() now returns base64-encoded images for LLM vision
  • Simplified function signature (removed over-engineered parameters)
  • Added self.screenshots_dir initialization
• Reorganized prompts structure
  • Created prompts/ folder for better organization
  • Synced all 5 prompt files with standalone browser-agent project
  • Updated PROMPT_PATH to prompts/agent.md

Installation

pip install --upgrade connectonion
# or
pip install connectonion==0.7.3

Breaking Changes

None

Contributors

Thanks to all contributors! 🎉

Full Changelog: v0.7.2...v0.7.3