Skip to content

feat: add vision core, feishu watchdog, and enhanced system prompts#193

Open
pandyzhou wants to merge 2 commits intolsdefine:mainfrom
pandyzhou:feat/vision-and-feishu
Open

feat: add vision core, feishu watchdog, and enhanced system prompts#193
pandyzhou wants to merge 2 commits intolsdefine:mainfrom
pandyzhou:feat/vision-and-feishu

Conversation

@pandyzhou
Copy link
Copy Markdown

Summary

This PR brings two related enhancements to the GenericAgent project:

1. Vision Core & Feishu Watchdog (075dab4)

  • vision_core.py (250 lines): A new ask_vision function for multimodal vision queries, with image encoding utilities (image_to_data_url, _encode_png_bytes, _build_user_message).
  • feishu_watchdog.py (51 lines): A watchdog process to monitor and auto-restart the Feishu/Lark bot integration, improving service reliability.
  • tests/test_vision_core.py (138 lines): Unit tests covering the vision core functionality.
  • requirements.txt: Project dependencies including Pillow, requests, etc.
  • Updated vision_sop.md and fsapp.py (connection retry logic, watchdog integration).

2. System Prompts Overhaul (f4b4c95)

  • Expanded sys_prompt.txt and sys_prompt_en.txt with new sections: Personality, Values, Interaction Style, and Coding Guidelines.
  • Minor adjustment in fsapp.py for default LLM selection.

Files Changed

Type File
New vision_core.py
New frontends/feishu_watchdog.py
New tests/test_vision_core.py
New requirements.txt
Modified assets/sys_prompt.txt
Modified assets/sys_prompt_en.txt
Modified frontends/fsapp.py
Modified memory/vision_sop.md

Stats: +786 / -12 across 8 files

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant