- Add call_vision method to all bot implementations (DashScope, Claude,
Gemini, ZhipuAI, MiniMax, Doubao, Moonshot, OpenAICompatibleBot)
using each vendor's native multimodal API format
- Remove call_with_tools/call_vision from Bot base class to fix MRO
shadowing issue with OpenAICompatibleBot mixin
- Refactor vision tool provider resolution: MainModel → other configured
models (auto-discovered) → OpenAI → LinkAI, with automatic fallback
- Return actual model name used in call_vision responses
- Sync config.json API keys to .env bidirectionally on startup
- Fix bot instance cache to detect bot_type/use_linkai config changes
- Add SSE reconnection support for web console
- Preserve image path hints in Gemini text for correct vision tool calls
- Update docs/tools/vision.mdx
Previously, files with extensions not in the known categories (image, document, video, audio) fell through to a fallback that returned ReplyType.TEXT, causing the file to never actually be sent to the user. Now the fallback uses ReplyType.FILE so all file types are delivered.
Made-with: Cursor
Browser tool enhancements:
- Navigate action now auto-includes snapshot result, saving one LLM round-trip
- Wait for networkidle + 800ms after navigation for SPA/JS-rendered pages
- Prompt guides agent to screenshot key results and ask user for login/CAPTCHA help
- Fixed playwright version pinned to 1.52.0; mirror fallback to official CDN on failure
Web console file/image support:
- SSE real-time push for images and files via on_event (file_to_send)
- Added /api/file endpoint to serve local files for web preview
- Frontend renders images in media-content container (survives delta/done overwrites)
- File attachment cards with download links; RFC 5987 encoding for non-ASCII filenames
Tool workspace fix:
- Inject workspace_dir as cwd into send and browser tools (previously only file tools)
- Screenshots now save to ~/cow/tmp/ instead of project directory
Separate DeepSeek from ChatGPTBot into its own module (models/deepseek/) with dedicated deepseek_api_key and deepseek_api_base config fields, avoiding config conflicts when switching between providers. Backward compatible with old users who configured DeepSeek via open_ai_api_key/open_ai_api_base through automatic fallback.
Made-with: Cursor
- Use LLM to summarize discarded context into concise daily memory entries
- Batch trim to half when exceeding max_turns/max_tokens, reducing flush frequency
- Run summarization asynchronously in background thread, no blocking on replies
- Add daily scheduled flush (23:55) as fallback for low-activity days
- Sync trimmed messages back to agent to keep context state consistent
- Web console Channels page: display active channels as config cards, support
save/connect/disconnect with real-time start/stop of channel processes
- Custom dropdown for channel selection (consistent with model selector style),
custom confirmation dialog for disconnect
- Fix channel stop: use sys.modules['__main__'] to access live ChannelManager
- Fix web request pending: move stop logic outside lock, set daemon_threads=True
- Fix reconnect: new asyncio event loop per startup, ctypes thread interrupt,
5s grace period before re-establishing remote connection
- Filter stale offline messages (>60s) pushed after reconnect
- Fix dark mode FOUC: apply theme in <head> before first paint, defer
transition-colors to post-init to avoid animated flash on load
- Fix Safari IME Enter bug: defer compositionend reset via setTimeout(0)
- Fix history scroll: use requestAnimationFrame before scrollChatToBottom
- Limit restore turns to min(6, max_turns//3) on restart
- Fix load_messages cutoff to start at turn boundary, preventing orphaned
tool_use/tool_result pairs from being sent to the LLM
- Merge all assistant messages within one user turn into a single bubble;
render tool_calls in history using same CSS as live SSE view
- Handle empty choices list in stream chunks
When numeric model names (e.g., '1') are used with vLLM and configured
in YAML without quotes, they are parsed as integers. This causes
AttributeError when calling startswith() method.
Changes:
- Add type checking for model_type
- Convert non-string model_type to string with warning log
- Prevents crash when using custom numeric model names
Fixes#2664
