Add SqliteSaver-based checkpointing so crashed analyses resume from the
last successful graph node instead of restarting from scratch.
- checkpointer.py: get_checkpointer(), thread_id(), has/clear_checkpoint()
- --checkpoint flag (default: off for backward compatibility)
- --clear-checkpoints flag to force fresh start
- Per-ticker SQLite DBs for parallel worker safety
- Logs 'Resuming from step N' vs 'Starting fresh'
- Clears checkpoint on successful completion (no stale state)
- Tests: crash resume + different date starts fresh
Add effort parameter (high/medium/low) for Claude 4.5+ and 4.6 models,
consistent with OpenAI reasoning_effort and Google thinking_level.
Also add content normalization for Anthropic responses.
InvestDebateState was missing bull_history, bear_history, judge_decision.
RiskDebateState was missing aggressive_history, conservative_history,
neutral_history, latest_speaker, judge_decision. This caused KeyError
in _log_state() and reflection, especially with edge-case config values.
Prevents UnicodeEncodeError on Windows where the default encoding
(cp1252/gbk) cannot handle Unicode characters in LLM output.
Closes#77, closes#114, closes#126, closes#215, closes#332
- Add StatsCallbackHandler for tracking LLM calls, tool calls, and tokens
- Integrate callbacks into TradingAgentsGraph and all LLM clients
- Dynamic agent/report counts based on selected analysts
- Fix report completion counting (tied to agent completion)
Clayton Brown
Yijia-Xiao
makk9
luohy15
mirza-samad-ahmed-baig
Edward Sun
maxer137