FinancialSituationMemory stored all learned lessons in RAM-only Python
lists with no persistence layer. Every process restart or new
TradingAgentsGraph() instance wiped all memory, making
reflect_and_remember() useless in practice — especially in server/API
deployments where a new graph is created per request.
Changes:
1. memory.py — add optional JSON file persistence:
- New config key `memory_persist_dir`: when set to a directory path,
each memory instance writes `<dir>/<name>.json` on every mutation
(add_situations / clear) and loads it on construction.
- When unset or None (the default), behaviour is identical to before
(RAM-only) — fully backward compatible.
- Atomic-ish writes via .tmp → rename to avoid corruption on crash.
- Graceful handling of corrupt / missing / partial JSON files.
- Tilde expansion (`~/...`) and automatic parent directory creation.
2. default_config.py — add `memory_persist_dir: None` to DEFAULT_CONFIG.
3. main.py — enable persistence in the example and improve
reflect_and_remember documentation comment.
4. tests/test_memory_persistence.py — 21 regression tests covering:
- RAM-only backward compatibility (5 tests)
- Persistence round-trip, incremental add, clear, BM25 rebuild,
JSON schema, Unicode (7 tests)
- Edge cases: corrupt JSON, missing keys, mismatched lengths,
nested directory creation, tilde expansion (5 tests)
- Multiple instances sharing same directory (1 test)
- Default config key existence (2 tests)
- Source audit: TradingAgentsGraph passes config to all 5 memories (1 test)
Closes#563
Apply review suggestions: use concise `or` pattern for API key
resolution, consolidate tests into parameterized subTest, move
import to module level per PEP 8.
GoogleClient now accepts the unified `api_key` parameter used by
OpenAI and Anthropic clients, mapping it to the provider-specific
`google_api_key` that ChatGoogleGenerativeAI expects. Legacy
`google_api_key` still works for backward compatibility.
Resolves TODO.md item #2 (inconsistent parameter handling).
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.
- Point requirements.txt to pyproject.toml as single source of truth
- Resolve welcome.txt path relative to module for CLI portability
- Include cli/static files in package build
- Extract shared normalize_content for OpenAI Responses API and
Gemini 3 list-format responses into base_client.py
- Update README install and CLI usage instructions
Enable use_responses_api for native OpenAI provider, which supports
reasoning_effort with function tools across all model families.
Removes the UnifiedChatOpenAI subclass workaround.
Closes#403
- Add http_client and http_async_client parameters to all LLM clients
- OpenAIClient, GoogleClient, AnthropicClient now support custom httpx clients
- Fixes SSL certificate verification errors on Windows Conda environments
- Users can now pass custom httpx.Client with verify=False or custom certs
Fixes#369
- OpenAI: add GPT-5.4, GPT-5.4 Pro; remove o-series and legacy GPT-4o
- Anthropic: add Claude Opus 4.6, Sonnet 4.6; remove legacy 4.1/4.0/3.x
- Google: add Gemini 3.1 Pro, 3.1 Flash Lite; remove deprecated
gemini-3-pro-preview and Gemini 2.0 series
- xAI: clean up model list to match current API
- Simplify UnifiedChatOpenAI GPT-5 temperature handling
- Add missing tradingagents/__init__.py (fixes pip install building)
Add _clean_dataframe() to normalize stock DataFrames before stockstats:
coerce invalid dates/prices, drop rows missing Close, fill price gaps.
Also add on_bad_lines="skip" to all cached CSV reads.
LLMs (especially smaller models) sometimes pass multiple indicator
names as a single comma-separated string instead of making separate
tool calls. Split and process each individually at the tool boundary.
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
d 🔹
voidborne-d
Yijia-Xiao
Yijia Xiao
CadeYu
javierdejesusda
阳虎
makk9