📸 Screenshots

Features

• 🧠 Brain with Memory — ChromaDB vector store retains trade experiences, semantic rules, and system rejections. Past outcomes are retrieved by similarity to current market conditions and injected into every LLM prompt.
• 📈 Vision AI Chart Analysis — Generates 4K PNG candlestick charts with indicators, sends them to a multimodal LLM (Gemini 3.5 Flash) for visual pattern recognition. Chart-pattern code was dropped because the AI reads charts better than hardcoded rules.
• 🔄 Reflection Engine — After every N closed trades, the system synthesizes best-practice rules, anti-patterns, and AI-mistake rules. These persist in vector memory and influence future decisions — the bot learns from its own outcomes.
• ✅ Claim Validation — Every LLM response is cross-checked against computed indicators. Reported trend strength is compared against actual ADX; pattern quality is replaced by a deterministic scorer. No blind trust in AI numeric claims.
• 📰 RAG News Engine — Aggregates crypto news from free RSS feeds (CoinDesk, CoinTelegraph, Decrypt, CryptoSlate) with optional Crawl4AI enrichment, plus fundamentals from DeFiLlama.
• 📊 Live Dashboard — FastAPI + WebSocket real-time UI at 0.0.0.0:8000 (or semanticsignal.qrak.org). Shows brain activity, last prompt/response, position state, performance stats, news, market data, and memory bank.
• 🛡️ Risk Pipeline — Pre-execution guard chain (symbol whitelist, max position size, cooldown) + dynamic SL/TP scaling with minimum 1.5 R:R enforced. Soft exits at candle close, hard exits at configurable intervals against live ticker price.
• 🔄 Multi-Provider AI Routing — Primary: Google Gemini 3.5 Flash (free tier). Fallback chain through OpenRouter and LM Studio. Chart vision support on every provider that allows it.
• 🧪 900+ Tests — Fully mocked test suite covering LLM output corruption, async races, rate-limit backoff, vector-DB boundaries, friction-reporting, and closed-loop feedback.

Architecture

flowchart TB
    subgraph Data["Data Sources"]
        EX["Exchanges (CCXT) → OHLCV + Order Book + Trade Flow"]
        NEWS["RSS Feeds + Crawl4AI"]
        FUND["CoinGecko + DeFiLlama + Alternative.me"]
    end
    subgraph Analysis["Analysis Engine"]
        TC["Technical Calculator<br/>40+ indicators"]
        PE["Pattern Engine<br/>Deterministic indicator patterns"]
        CG["Chart Generator<br/>4K PNG with SMA/RSI/Volume"]
        RAG["RAG Engine<br/>News relevance scoring"]
    end
    subgraph Brain["🧠 Brain Layer"]
        VM["Vector Memory<br/>ChromaDB (3 collections)"]
        REFL["Reflection Engine<br/>Rules from closed trades"]
        CTX["Context Builder<br/>Similarity retrieval + confidence calibration"]
    end
    subgraph Execution["Paper Execution"]
        RP["Risk Manager<br/>SL/TP, sizing, R:R"]
        GP["Guard Pipeline<br/>Symbol → Size → Cooldown"]
        STRAT["Trading Strategy<br/>ExitMonitor + PositionStatusMonitor"]
    end
    Data --> Analysis
    Analysis --> Brain
    Brain --> AI["AI Provider<br/>(Gemini / OpenRouter / LM Studio)"]
    AI --> RP --> GP --> STRAT
    STRAT -.->|Closed trade feedback| Brain

Key Files

Testing

# Full suite (900+ tests)
pytest tests/ -q

# Focused
pytest tests/test_ticker_retry.py tests/test_brain_integration.py -q

# Linting
ruff check src tests start.py

Configuration

Key settings in config/config.ini:

Required API keys in keys.env:

Roadmap

• Multiple Trading Agent Personalities — Conservative, aggressive, contrarian, trend-following strategists (aspirational)
• Multi-Model Consensus — "Council of Models" architecture for collective decision-making (aspirational)
• Live Trading — Exchange order execution layer (plan at .ai/plans/real_trading_implementation_plan.md)
• Admin Dashboard — Web GUI for bot configuration (replaces manual config.ini editing) (plan at .ai/plans/admin-dashboard-plan.md)

Disclaimer

NOT FINANCIAL ADVICE. This software is experimental and in BETA. Paper-trading only — real exchange order execution is not implemented. No warranty provided. Use at your own risk.

License

MIT