2145b04318
fix: graceful LLM 404 handling + per-tier fallback model config ( #108 )
...
* fix: per-tier fallback LLM for provider 404/policy errors
- tool_runner: catch status_code==404 from chain.invoke(), re-raise as
RuntimeError with actionable message (OpenRouter privacy URL + env var hint)
- langgraph_engine: wrap astream_events in try/except, detect policy errors
and re-raise with model/provider context
- langgraph_engine: _run_one_ticker distinguishes policy 404s (logger.error,
no traceback) from real bugs (logger.exception with traceback); if fallback
is configured, rebuilds pipeline with fallback model tier and retries
- langgraph_engine: add _is_policy_error() and _build_fallback_config() helpers
- default_config: add quick/mid/deep_think_fallback_llm + _provider keys
(TRADINGAGENTS_QUICK_THINK_FALLBACK_LLM etc.)
- .env.example: document new fallback env vars
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* docs: ADR 017 LLM policy fallback, correct ADR 016 findings, update CLAUDE.md
- docs/agent/decisions/017: add ADR for per-tier LLM fallback design decision
- docs/agent/decisions/016: correct 3 inaccurate review findings — list_pm_decisions
ObjectId projection, created_at datetime type, and base_dir pointer handling are
all already correctly implemented in PR#106
- CLAUDE.md: add Per-Tier Fallback LLM section and _is_policy_error critical pattern
- CURRENT_STATE.md: update milestone and recent progress for PR#106/107/108 merges
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
---------
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-03-25 11:19:08 +01:00
9c9cc8c0b6
fix: address all PR#106 review findings (ADR 016) ( #106 )
...
* Initial plan
* feat: add observability logging - run event persistence and enriched tool events
- Integrate RunLogger into LangGraphEngine for JSONL event persistence
- Add _start_run_logger/_finish_run_logger lifecycle in all run methods
- Enrich tool events with service, status, and error fields
- Add _TOOL_SERVICE_MAP for tool-to-service name resolution
- Frontend: color error events in red, show service badges
- Frontend: display graceful_skip status with orange indicators
- Frontend: add error tab and service info to EventDetail/EventDetailModal
- Add 11 unit tests for new observability features
Co-authored-by: aguzererler <6199053+aguzererler@users.noreply.github.com>
Agent-Logs-Url: https://github.com/aguzererler/TradingAgents/sessions/477a0676-af7b-48ff-8a3d-567e943323cf
* refactor: address code review - extract graceful keywords constant, fix imports
- Move get_daily_dir import to top-level (remove inline aliases)
- Extract _GRACEFUL_SKIP_KEYWORDS as module-level constant
- Update test patches to match top-level import location
Co-authored-by: aguzererler <6199053+aguzererler@users.noreply.github.com>
Agent-Logs-Url: https://github.com/aguzererler/TradingAgents/sessions/477a0676-af7b-48ff-8a3d-567e943323cf
* feat: add run_id to report paths, MongoDB report store, store factory, and reflexion memory
- report_paths.py: All path helpers accept optional run_id for run-scoped dirs
- report_store.py: ReportStore supports run_id + latest.json pointer mechanism
- mongo_report_store.py: MongoDB-backed store with same interface (no overwrites)
- store_factory.py: Factory returns MongoDB or filesystem store based on config
- memory/reflexion.py: Reflexion memory for learning from past decisions
- langgraph_engine.py: Uses store factory + run_id for all run methods
- Fix save_holding_reviews bug (was save_holding_reviews, now save_holding_review)
- default_config.py: Add mongo_uri and mongo_db config keys
- pyproject.toml: Add pymongo>=4.12.1 dependency
- .env.example: Document TRADINGAGENTS_MONGO_URI and TRADINGAGENTS_MONGO_DB
Co-authored-by: aguzererler <6199053+aguzererler@users.noreply.github.com>
Agent-Logs-Url: https://github.com/aguzererler/TradingAgents/sessions/16e673ea-40a1-40a0-8e77-f8cd08c1a716
* fix: clean up reflexion record_outcome (remove broken update_one with sort)
Also update runs.py reset endpoint to use store factory, fix tests,
add ADR 015, update CURRENT_STATE.md
Co-authored-by: aguzererler <6199053+aguzererler@users.noreply.github.com>
Agent-Logs-Url: https://github.com/aguzererler/TradingAgents/sessions/16e673ea-40a1-40a0-8e77-f8cd08c1a716
* fix: address all PR#106 review findings (ADR 016)
- Fix save_holding_review: iterate per-ticker instead of passing
portfolio_id as ticker (Finding 13)
- Fix RunLogger context: replace threading.local with contextvars
for asyncio task isolation (Finding 3)
- Fix list_pm_decisions: add _id:0 projection to exclude ObjectId (Finding 6)
- Fix ReflexionMemory: native datetime for MongoDB, ISO string for
local JSON fallback (Finding 7)
- Fix latest pointer: write/read_latest_pointer accept base_dir
parameter, ReportStore passes _base_dir (Finding 12)
- Wire RunLogger callback into all astream_events calls (Finding 1)
- Call ensure_indexes in MongoReportStore.__init__ (Finding 11)
- Create ADR 016 documenting all 13 findings and resolutions
- Add 14 targeted tests covering all 7 fixes
- All 886 tests pass (872 existing + 14 new)
Co-authored-by: aguzererler <6199053+aguzererler@users.noreply.github.com>
Agent-Logs-Url: https://github.com/aguzererler/TradingAgents/sessions/e52cdd2f-efae-4d2a-a56f-903d909b3342
* chore: remove unused imports in tests, remove redundant ensure_indexes call in factory
Co-authored-by: aguzererler <6199053+aguzererler@users.noreply.github.com>
Agent-Logs-Url: https://github.com/aguzererler/TradingAgents/sessions/e52cdd2f-efae-4d2a-a56f-903d909b3342
* docs: update ADR 016 — mark Finding 2 resolved, update context docs for contextvars
Co-authored-by: aguzererler <6199053+aguzererler@users.noreply.github.com>
Agent-Logs-Url: https://github.com/aguzererler/TradingAgents/sessions/ce9e2400-a60d-4a6b-896b-1b34ec786bed
---------
Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: aguzererler <6199053+aguzererler@users.noreply.github.com>
2026-03-25 11:14:23 +01:00
cd655cdae5
feat: concurrent per-ticker pipelines in auto mode, controlled by TRADINGAGENTS_MAX_CONCURRENT_PIPELINES ( #103 )
...
* Initial plan
* feat: concurrent per-ticker pipelines in auto mode, configurable via TRADINGAGENTS_MAX_CONCURRENT_PIPELINES
Co-authored-by: aguzererler <6199053+aguzererler@users.noreply.github.com>
Agent-Logs-Url: https://github.com/aguzererler/TradingAgents/sessions/464f2038-3a60-4d86-9fe8-4e1cc6943174
---------
Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: aguzererler <6199053+aguzererler@users.noreply.github.com>
Co-authored-by: ahmet guzererler <guzererler@gmail.com>
2026-03-24 18:08:06 +01:00
4c186a55e8
merge: sync with upstream TauricResearch/TradingAgents v0.2.2
...
Merges upstream/main into our fork, bringing in:
- Security: Remove chainlit (CVE-2026-22218), patch LangGrinch vulnerability
- Bug fixes: debate state init, UTF-8 encoding, stock data parsing, debate round config
- Features: OpenAI Responses API, five-tier rating scale, Anthropic effort support,
exchange-qualified tickers, yfinance retry with exponential backoff
- Refactor: risk_manager renamed to portfolio_manager
Conflicts resolved preserving our fork's:
- Per-tier LLM config (quick/mid/deep think with provider overrides)
- Separate tool files (fundamental_data_tools, core_stock_tools, etc.)
- Custom tools (get_ttm_analysis, get_peer_comparison, get_sector_relative, get_macro_regime)
- Dynamic Ollama model fetching
- Enhanced fundamentals analyst prompt with TTM analysis
- Hardened stockstats/yfinance data pipeline (_load_or_fetch_ohlcv)
- AgentOS observability layer, scanner pipeline, portfolio management
Tests: 727 passed, 14 skipped
2026-03-23 12:17:25 +00:00
bd9b1e5efa
feat: add Anthropic effort level support for Claude models
...
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.
2026-03-22 21:57:05 +00:00
4a3aa5ca3c
refactor: centralize env loading in default_config.py; fix .env.example
...
Co-authored-by: aguzererler <6199053+aguzererler@users.noreply.github.com>
Agent-Logs-Url: https://github.com/aguzererler/TradingAgents/sessions/b62cf290-b389-45bf-b644-08b2360a1f76
2026-03-21 22:33:06 +00:00
179e55f264
fix(tests): complete PR #26 — enforce socket isolation in unit tier and add test suite reference doc ( #30 )
...
* Initial plan
* fix(tests): complete PR #26 — move integration tests to tests/integration/ and fix 4 failing unit tests
Co-authored-by: aguzererler <6199053+aguzererler@users.noreply.github.com>
* docs: create docs/testing.md — comprehensive test suite reference with 5 Mermaid flowcharts
Co-authored-by: aguzererler <6199053+aguzererler@users.noreply.github.com>
---------
Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com>
Co-authored-by: aguzererler <6199053+aguzererler@users.noreply.github.com>
2026-03-19 17:41:25 +01:00
a90f14c086
feat: unified report paths, structured observability logging, and memory system update ( #22 )
...
* gitignore
* feat: unify report paths under reports/daily/{date}/ hierarchy
All generated artifacts now land under a single reports/ tree:
- reports/daily/{date}/market/ for scan results (was results/macro_scan/)
- reports/daily/{date}/{TICKER}/ for per-ticker analysis (was reports/{TICKER}_{timestamp}/)
- reports/daily/{date}/{TICKER}/eval/ for eval logs (was eval_results/{TICKER}/...)
Adds tradingagents/report_paths.py with centralized path helpers used by
CLI commands, trading graph, and pipeline.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
* feat: structured observability logging for LLM, tool, and vendor calls
Add RunLogger (tradingagents/observability.py) that emits JSON-lines events
for every LLM call (model, agent, tokens in/out, latency), tool invocation
(tool name, args, success, latency), data vendor call (method, vendor,
success/failure, latency), and report save.
Integration points:
- route_to_vendor: log_vendor_call() on every try/catch
- run_tool_loop: log_tool_call() on every tool invoke
- ScannerGraph: new callbacks param, passes RunLogger.callback to all LLM tiers
- pipeline/macro_bridge: picks up RunLogger from thread-local, passes to TradingAgentsGraph
- cli/main.py: one RunLogger per command (analyze/scan/pipeline), write_log()
at end, summary line printed to console
Log files co-located with reports:
reports/daily/{date}/{TICKER}/run_log.jsonl (analyze)
reports/daily/{date}/market/run_log.jsonl (scan)
reports/daily/{date}/run_log.jsonl (pipeline)
Also fix test_long_response_no_nudge: update "A"*600 → "A"*2100 to match
MIN_REPORT_LENGTH=2000 threshold set in an earlier commit.
Update memory system context files (ARCHITECTURE, COMPONENTS, CONVENTIONS,
GLOSSARY, CURRENT_STATE) to document observability and report path systems.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
---------
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-19 09:06:40 +01:00
26cd4c8b78
feat: Finnhub integration layer, 141 tests, and vendor evaluation report ( #16 )
...
* feat: add Finnhub integration layer, tests, and evaluation report
Adds a complete Finnhub data vendor integration as a supplementary
source alongside Alpha Vantage — zero changes to existing functionality.
New dataflow modules:
- finnhub_common.py: exception hierarchy, thread-safe rate limiter (60/min), _make_api_request
- finnhub_stock.py: get_stock_candles, get_quote
- finnhub_fundamentals.py: get_company_profile, get_financial_statements, get_basic_financials
- finnhub_news.py: get_company_news, get_market_news, get_insider_transactions
- finnhub_scanner.py: market movers (S&P 500 basket workaround), indices, sectors, topic news
- finnhub_indicators.py: SMA, EMA, MACD, RSI, BBANDS, ATR via /indicator endpoint
- finnhub.py: facade re-exporting all public functions
New tests:
- test_finnhub_integration.py: 100 offline (mocked HTTP) tests — all passing
- test_finnhub_live_integration.py: 41 live integration tests — skip gracefully when FINNHUB_API_KEY unset
Evaluation report (docs/finnhub_evaluation.md):
- Full coverage matrix vs Alpha Vantage across 5 data categories
- Free tier viability analysis (60 calls/min)
- Unique capabilities: earnings calendar, economic calendar, XBRL as-filed filings
- Recommendation: add as supplementary vendor for calendar data only
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* test: mark paid-tier Finnhub endpoints; update evaluation with live results
Live testing with free-tier key confirmed:
- /quote, /stock/profile2, /stock/metric, /company-news, /news,
/stock/insider-transactions → all free tier (27 live tests PASS)
- /stock/candle, /financials-reported, /indicator → paid tier HTTP 403
(14 tests now properly skipped with @pytest.mark.paid_tier)
Also:
- Register 'integration' and 'paid_tier' markers in pyproject.toml
- Update docs/finnhub_evaluation.md with confirmed endpoint availability table
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
* feat: wire Finnhub into routing layer — insider txns, calendars, fallback
Changes:
- interface.py: Finnhub added as third vendor (alongside yfinance + AV)
- get_insider_transactions: Finnhub primary (free, + MSPR bonus signal)
- get_market_indices/sector_performance/topic_news: Finnhub added as option
- Fallback catch extended: (AlphaVantageError, FinnhubError, ConnectionError, TimeoutError)
- New calendar_data category with get_earnings_calendar + get_economic_calendar
- finnhub_scanner.py: added get_earnings_calendar_finnhub, get_economic_calendar_finnhub
(FOMC/CPI/NFP/GDP events + earnings beats — unique, not in AV at any tier)
- finnhub.py: re-exports new calendar functions
- scanner_tools.py: @tool wrappers for get_earnings_calendar, get_economic_calendar
- default_config.py: tool_vendors["get_insider_transactions"]="finnhub",
calendar_data vendor category defaulting to "finnhub"
- .env.example: FINNHUB_API_KEY documented
- docs/agent/decisions/010-finnhub-vendor-integration.md: ADR for this decision
All 173 offline tests pass. ADR 002 constraints respected throughout.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
---------
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-03-18 11:28:43 +01:00
2193ff3fa1
fix: allow .env variables to override DEFAULT_CONFIG values
...
Merged origin/main and resolved all 8 conflicting files:
- CLAUDE.md: merged MISTAKES.md ref + Project Tracking section + env override docs
- cli/main.py: kept top-level json import, kept try/except in run_pipeline
- tool_runner.py: kept descriptive comments for MAX_TOOL_ROUNDS
- alpha_vantage_common.py: kept thread-safe rate limiter, robust error handling
- interface.py: kept broader exception catch (AlphaVantageError + ConnectionError + TimeoutError)
- default_config.py: kept _env()/_env_int() env var overrides with load_dotenv() at module level
- scanner_graph.py: kept debug mode fix (stream for debug, invoke for result)
- macro_bridge.py: kept get_running_loop() over deprecated get_event_loop()
Co-authored-by: aguzererler <6199053+aguzererler@users.noreply.github.com>
2026-03-17 14:25:10 +00:00
61668fed6b
feat: Complete 3-phase LLM scanner pipeline with inline tool execution
...
The scanner pipeline now runs end-to-end: Phase 1 (geopolitical, market
movers, sector scanners in parallel via Ollama), Phase 2 (industry deep
dive), Phase 3 (macro synthesis via OpenRouter/DeepSeek R1).
Key changes:
- Add tool_runner.py with run_tool_loop() for inline tool execution in
scanner agents (scanner graph has no ToolNode, unlike trading graph)
- Fix vendor fallback: catch AlphaVantageError base class, raise on
total failure instead of embedding errors in return values
- Rewrite yfinance sector perf to use SPDR ETF proxies (Sector.overview
has no performance data)
- Fix Ollama remote host support in openai_client.py
- Add LangGraph state reducers for parallel fan-out writes
- Add --date CLI flag for non-interactive scanner invocation
- Fix .env loading to find keys from both CWD and project root
- Add hybrid LLM config (per-tier provider/backend_url)
- Add project tracking: DECISIONS.md, PROGRESS.md, MISTAKES.md
- Add 9 new test files covering exceptions, fallback, and routing
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-03-17 08:41:40 +01:00
6242af3b99
feat: Add Global Macro Scanner feature
...
- Added market-wide analysis capabilities (movers, indices, sectors, industries, topic news)
- Implemented yfinance and Alpha Vantage data fetching modules
- Added LangChain tools for scanner functions
- Created scanner state definitions and graph components
- Integrated scan command into CLI
- Added configuration for scanner_data vendor routing
- Included test files for scanner components
This implements a new feature for global macro scanning to identify market-wide trends and opportunities.
2026-03-14 22:22:13 +01:00
c471db80e1
feat(cli): add three-tier LLM provider setup with dynamic Ollama model listing
...
- Add quick/mid/deep thinking tiers, each with independent provider selection
- Fetch available Ollama models dynamically via /api/tags instead of hardcoded list
- Add mid-thinking agent selection (select_mid_thinking_agent)
- Support provider-specific thinking config (Gemini thinking level, OpenAI reasoning effort)
- Update default_config and trading_graph to wire three-tier LLM setup
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-03-13 13:20:50 +01:00
3d040f8da4
feat: add yfinance support to accommodate community request for stability and quota
2026-02-03 22:27:20 +00:00
d4dadb82fc
feat: add multi-provider LLM support with thinking configurations
...
Models added:
- OpenAI: GPT-5.2, GPT-5.1, GPT-5, GPT-5 Mini, GPT-5 Nano, GPT-4.1
- Anthropic: Claude Opus 4.5/4.1, Claude Sonnet 4.5/4, Claude Haiku 4.5
- Google: Gemini 3 Pro/Flash, Gemini 2.5 Flash/Flash Lite
- xAI: Grok 4, Grok 4.1 Fast (Reasoning/Non-Reasoning)
Configs updated:
- Add unified thinking_level for Gemini (maps to thinking_level for Gemini 3,
thinking_budget for Gemini 2.5; handles Pro's lack of "minimal" support)
- Add OpenAI reasoning_effort configuration
- Add NormalizedChatGoogleGenerativeAI for consistent response handling
Fixes:
- Fix Bull/Bear researcher display truncation
- Replace ChromaDB with BM25 for memory retrieval
2026-02-03 22:27:20 +00:00
b01051b9f4
Switch default data vendor
...
🤖 Generated with [Claude Code](https://claude.com/claude-code )
Co-Authored-By: Claude <noreply@anthropic.com>
2025-09-30 12:43:27 +08:00
0ab323c2c6
Add Alpha Vantage API integration as primary data provider
...
- Replace FinnHub with Alpha Vantage API in README documentation
- Implement comprehensive Alpha Vantage modules:
- Stock data (daily OHLCV with date filtering)
- Technical indicators (SMA, EMA, MACD, RSI, Bollinger Bands, ATR)
- Fundamental data (overview, balance sheet, cashflow, income statement)
- News and sentiment data with insider transactions
- Update news analyst tools to use ticker-based news search
- Integrate Alpha Vantage vendor methods into interface routing
- Maintain backward compatibility with existing vendor system
🤖 Generated with [Claude Code](https://claude.ai/code )
Co-Authored-By: Claude <noreply@anthropic.com>
2025-09-26 22:57:50 +08:00
a6734d71bc
WIP
2025-09-26 16:17:50 +08:00
ee3d499894
Merge branch 'TauricResearch:main' into save_results
2025-06-25 08:43:19 +09:00
da84ef43aa
main works, cli bugs
2025-06-15 22:20:59 -07:00
9647359246
save reports & logs under results_dir
2025-06-12 11:25:07 +09:00
99789f9cd1
Add support for other backends, such as OpenRouter and olama
...
This aims to offer alternative OpenAI capable api's.
This offers people to experiment with running the application locally
2025-06-11 14:19:25 +02:00
cc97cb6d5d
chore(release): v0.1.0 – initial public release of TradingAgents
2025-06-05 04:27:57 -07:00
ahmet guzererler
Copilot
Yijia-Xiao
luohy15
Huijae Lee
Edward Sun
maxer137