Commit Graph

26 Commits

Author SHA1 Message Date
Andrew Kaszubski 4f6f7c1c14 feat(memory): add memory integration for agent prompts - Fixes #21 2025-12-26 20:40:21 +11:00
Andrew Kaszubski 5a0606b59f feat(graph): integrate new analysts into workflow - Fixes #17 2025-12-26 20:11:28 +11:00
Andrew Kaszubski a17fc1f029 feat(agents): add Position Sizing Manager with Kelly and risk parity - Fixes #16
Implements Position Sizing Manager for optimal position sizing calculations.
The manager supports multiple sizing methodologies to accommodate different
trading styles and risk tolerances.

Sizing Methods:
- Kelly Criterion (full, half, quarter) for edge-based optimal sizing
- ATR-based sizing for volatility-adjusted position sizing
- Risk Parity allocation for balanced portfolio risk
- Volatility targeting for consistent risk contribution
- Fixed fractional for controlled percentage risk

Key Features:
- Multiple sizing method comparison for best fit selection
- Risk level presets (conservative, moderate, aggressive)
- Position constraints and drawdown limits
- Stop loss level recommendations based on ATR
- Win/loss ratio analysis for Kelly calculations

Tools:
- calculate_kelly_position_size: Edge-based optimal sizing
- calculate_atr_position_size: Volatility-adjusted position sizing
- calculate_risk_parity_allocation: Multi-asset balanced risk allocation
- calculate_volatility_target_size: Target volatility-based sizing
- get_position_sizing_recommendation: Comprehensive multi-method comparison

Enums:
- SizingMethod: 8 different sizing approaches
- RiskLevel: Conservative, moderate, aggressive presets

Tests: 52 unit tests covering Kelly calculations, ATR sizing, risk parity
weights, volatility targeting, constraints, and integration workflows.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
2025-12-26 17:39:20 +11:00
Andrew Kaszubski b0140a82b3 feat(agents): add Correlation Analyst for cross-asset analysis - Fixes #15
Implements Correlation Analyst agent for cross-asset correlation analysis
and sector rotation detection. The agent analyzes inter-market relationships
to identify diversification opportunities and sector leadership changes.

Features:
- Cross-asset correlation analysis (stocks vs bonds, gold, oil, dollar)
- Sector rotation analysis across all 11 S&P 500 sectors
- Rolling correlation trend tracking with breakdown detection
- Relative strength calculation for sector leadership
- Economic cycle phase identification (early/mid/late cycle, recession)
- Cycle-based sector recommendations

Tools:
- get_cross_asset_correlation_analysis: Multi-asset correlation with regime interpretation
- get_sector_rotation_analysis: Sector rankings, leadership, rotation signals
- get_correlation_matrix: Pairwise correlation matrix for portfolio construction
- get_rolling_correlation_trend: Time-series correlation dynamics

Enums:
- CorrelationStrength: 9-level classification from very strong negative to very strong positive
- SectorPhase: Economic cycle phases for rotation timing
- SectorLeadership: Leading, lagging, improving, weakening classifications

Tests: 59 unit tests covering correlation calculations, classification logic,
sector rotation, breakdown detection, and integration scenarios.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
2025-12-26 17:30:53 +11:00
Andrew Kaszubski bdff87a571 feat(agents): add Macro Analyst with FRED interpretation - Fixes #14
Implements Macro Analyst agent for economic regime detection and FRED data
interpretation. The agent analyzes macroeconomic indicators to identify
market regimes and provide investment implications.

Features:
- Economic regime detection (expansion, contraction, goldilocks, stagflation)
- Yield curve analysis (normal, inverted, flat, steep)
- Monetary policy stance classification (hawkish, dovish, neutral, emergency)
- Inflation regime tracking (deflation, low, target, elevated, high)
- Recession probability calculation based on yield curve inversions
- Real rate analysis for policy restrictiveness

Tools:
- get_economic_regime_analysis: GDP, unemployment, inflation regime detection
- get_yield_curve_analysis: 2Y-10Y spread analysis with recession probability
- get_monetary_policy_analysis: Fed policy stance interpretation
- get_inflation_regime_analysis: CPI/PCE trajectory and asset implications

Tests: 57 unit tests covering all classification logic, edge cases, and
integration scenarios.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
2025-12-26 17:23:30 +11:00
Andrew Kaszubski 8522b4bd53 feat(agents): add Momentum Analyst with multi-TF analysis - Fixes #13
Implements specialized Momentum Analyst agent with:
- Multi-timeframe ROC (Rate of Change) analysis across periods
- ADX (Average Directional Index) trend strength measurement
- RSI momentum divergence detection for reversal signals
- create_momentum_analyst factory for LangChain integration

Tests: 47 unit tests covering ROC, ADX, RSI, divergence detection

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
2025-12-26 17:12:01 +11:00
Andrew Kaszubski edae1ab2cc feat(llm): add DeepSeek provider and HuggingFace embedding fallback (Issue #41) 2025-12-26 11:07:48 +11:00
Andrew Kaszubski 9ee81be48b fix(memory): use get_or_create_collection for idempotent ChromaDB init - Fixes #30 2025-12-26 09:20:04 +11:00
Andrew Kaszubski 5443aaa209 feat(llm): add OpenRouter API support with proper headers and API key handling
- Add explicit OPENROUTER_API_KEY environment variable handling
- Add HTTP-Referer and X-Title headers for OpenRouter attribution
- Fix case sensitivity for provider names (ollama now case-insensitive)
- Add embedding fallback to OpenAI when using OpenRouter (since OpenRouter lacks embedding API)
- Add comprehensive test suite (30 tests) for OpenRouter integration
- Update README.md and PROJECT.md with OpenRouter configuration docs
- Add CHANGELOG.md documenting the changes

Patterns borrowed from ~/.claude/lib/genai_validate.py for multi-provider support.

Closes #1

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
2025-12-25 15:51:25 +11:00
Andrew Kaszubski 0bd3741e8a feat: Add Anthropic LLM support and fix data vendor compatibility
- Switch to Anthropic Claude as LLM provider in main.py

- Add get_google_global_news() wrapper for proper signature matching

- Add get_fundamentals() to yfinance vendor

- Register yfinance and google as vendors for fundamentals and global news

- Make memory system fail gracefully when OpenAI embeddings unavailable
2025-12-25 15:23:28 +11:00
luohy15 86bc0e793f minor fix 2025-09-27 00:04:59 +08:00
luohy15 8b04ec307f minor fix 2025-09-26 23:25:33 +08:00
luohy15 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
luohy15 a6734d71bc WIP 2025-09-26 16:17:50 +08:00
mirza-samad-ahmed-baig f704828f89 Fix: Prevent infinite loops, enable reflection, and improve logging 2025-07-03 17:43:40 +05:00
Max Wong 43aa9c5d09
Local Ollama (#53)
- Fix typo 'Start' 'End'
- Add llama3.1 selection
- Use 'quick_think_llm' model instead of hard-coding GPT
2025-06-26 00:27:01 -04:00
Yijia Xiao 26c5ba5a78
Revert "Docker support and Ollama support (#47)" (#57)
This reverts commit 78ea029a0b.
2025-06-26 00:07:58 -04:00
Geeta Chauhan 78ea029a0b
Docker support and Ollama support (#47)
- Added support for running CLI and Ollama server via Docker
- Introduced tests for local embeddings model and standalone Docker setup
- Enabled conditional Ollama server launch via LLM_PROVIDER
2025-06-25 23:57:05 -04:00
Yijia Xiao b575bd0941
Merge pull request #52 from TauricResearch/dev
Merge dev into main. Add support for Anthropic and OpenRouter.
2025-06-23 20:38:14 -04:00
Edward Sun 52284ce13c fixed anthropic support. Anthropic has different format of response when it has tool calls. Explicit handling added 2025-06-21 12:51:34 -07:00
Yijia Xiao c2f897fc67
Merge pull request #43 from AtharvSabde/patch-1
fundamentals_analyst.py (spelling mistake in instruction: Makrdown -> Markdown)
2025-06-19 15:05:08 -04:00
Atharv Sabde 2af7ef3d79
fundamentals_analyst.py(spelling mistake.markdown) 2025-06-19 21:48:16 +05:30
Edward Sun 7eaf4d995f update clear msg bc anthropic needs at least 1 msg in chat call 2025-06-15 23:14:47 -07:00
Edward Sun da84ef43aa main works, cli bugs 2025-06-15 22:20:59 -07:00
maxer137 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
Yijia-Xiao cc97cb6d5d chore(release): v0.1.0 – initial public release of TradingAgents 2025-06-05 04:27:57 -07:00