Prototype implementation that allows:
1. Normal conversation with the AI
2. Tool-calling to update trading strategies
Created new ConversationalAgent that uses CrewAI with tools:
- get_current_strategy: Check current bot strategy
- update_trading_strategy: Update bot's trading configuration
The agent can now respond to questions like 'What is this?' without
forcing JSON output, and can update strategies when user provides
specific parameters.
Refs #51
Changes:
1. Add 30-second timeout for chat API requests using AbortController
2. User's message now shows immediately before API response (already done in previous PR)
3. Differentiate between timeout errors and other errors in error messages
4. API client now accepts optional signal parameter for abort support
Previously, only the assistant's response was added to the frontend store.
Now both user and assistant messages are stored, so the conversation
displays correctly in the chat interface.
- Use CrewAI's LLM class directly with api_base parameter instead of custom subclass
- Remove broken MiniMaxLLM inheritance from LLM
- Update agent creation to use LLM(model, api_key, api_base) pattern
The issue was that inheriting from CrewAI's LLM class caused the api_base
to be set to None. Now we use CrewAI's LLM directly with the correct parameters.
Fixes#43
Changes:
1. Updated API endpoint from api.minimax.chat to api.minimax.io
2. Changed default model from MiniMax-Text-01 to MiniMax-M2.7
(MiniMax-Text-01 is not available for all API key plans)
3. Updated .env.example with correct default model
MiniMax API docs: https://platform.minimax.io/docs/api-reference/text-anthropic-apiFixes#43
After successful registration, the backend now returns an access token
(along with token_type) so the frontend can:
- Store the token in localStorage
- Fetch the user profile
- Redirect to dashboard
Fixes#41
bcrypt 5.0.0 is incompatible with passlib 1.7.x - passlib tries to
access bcrypt.__about__.__version__ which was removed in bcrypt 5.x.
Constrain bcrypt to >=4.0,<5.0 to maintain compatibility.
- Fix relative import path in crew.py (from ..core to ...core)
- Update __init__.py exports to match actual class names
- Remove incorrect CrewAgent and LLMConnector exports
Errors during price fetching are now logged and stored in an errors list,
allowing users to see error count/warnings in simulation results.
Acceptance Criteria:
- [x] Errors are logged (not silently swallowed)
- [x] User can see error count/warnings in simulation results
- [x] Simulation completes even if some price fetches fail (graceful degradation)
LLM was outputting nested params structure but engines expect flat fields.
This caused backtesting and simulation to never trigger any trades.
Changes:
- llm_connector.py: Update prompt to output flat condition structure
- crew.py: Update StrategyValidator to validate flat structure
- crew.py: Update StrategyExplainer to read flat structure
Fixes#25
