Skip to main content

The Trader-Agent Model

The agent handles everything that scales: scanning hundreds of markets, computing technical indicators, pulling positioning, sentiment and on-chain data, detecting behavioral patterns, logging snapshots, and generating diff summaries. None of this requires language model inference at scan time – it’s all conventional code against aggregated and synthesized cached data streams, running continuously at negligible cost. LLM reasoning is reserved for the filtered shortlist and direct trader queries, where reasoning actually adds value. The trader provides what does not scale: a thesis explaining why a trade makes sense right now, corrections when the agent’s setup misfires due to nuanced sentiment shifts, context for events not yet in the quantitative data, and a closing reflection on what the trade revealed.

The Trade Flow

1. Trader Asks, Agent Scans

Most trades start with a question. A trader asks about a specific pair, a market condition, or a thesis they want pressure-tested. The agent pulls the relevant intelligence layers, compresses them, and responds in plain language. Separately, the agent continuously monitors 200+ markets on Hyperliquid. A pre-filter pipeline computes technical indicators locally before invoking the language model, reducing inference costs by roughly 95% – only a filtered shortlist of 5-10 candidates requires LLM analysis. Whether the trader initiates or the agent surfaces something on its own, the next step is the same.

2. Agent Proposes

When a conversation or a scan produces a setup that meets the agent’s criteria, it presents a structured trade proposal to the trader:
  • Asset pair and direction (long/short)
  • Entry price, stop loss, take profit
  • Leverage and position size
  • Risk as percentage of equity
  • Reasoning – which data layers support the thesis – Portfolio context – recent trade history and any open positions that may compound exposure

3. Trader Decides

The trader reviews the proposal and responds:
  • Approve – Execute as proposed
  • Modify – Adjust parameters (move the stop, reduce size, change target)
  • Reject – Pass on this setup
  • Discuss – Ask the agent questions before deciding
Every interaction is logged. Trader overrides and corrections are the most valuable data in the lesson — they capture where human judgment diverges from machine analysis.

4. Lesson Captured

At position open, the agent captures a snapshot of market state across 15 categories (140 fields). At position close, it captures an identical snapshot and computes the difference. What changed in volume, funding rates, open interest, whale cohort positioning, sentiment, and smart money flows between entry and exit? This diff, combined with the trade outcome and the trader’s reasoning, produces a structured lesson.

5. Intelligence Compounds

Lessons sync to the central library. Pattern discovery tooling surfaces candidate patterns through feature importance analysis, unsupervised clustering across market regimes, and statistical validation with multiple-comparison correction. A curation team validates these candidates through backtesting, confirms robustness, and encodes proven patterns into shared skill files that all agents receive. When a skill file improves, every agent in the network improves simultaneously.

Trader Feedback Channels

Trader feedback enters the system through five channels of increasing friction and value:
ChannelFrictionValueDescription
Trade decisionsLowMediumApprove or reject setups through the conversational interface
CorrectionsLowHighOverride agent behavior (move stops, adjust size) - logged automatically
Context messagesMediumHighShare information the agent cannot observe
Weekly reviewsHighVery HighStructured reflections surfacing calendar-aware risk factors and broader narratives
Passive signals are also captured without trader effort: response latency, override frequency, activity patterns, and which agent recommendations are consistently accepted or rejected. Across thousands of traders, these behavioral signals reveal collective preferences that no individual trader articulates explicitly.

Bring Your Own LLM

Traders connect their own API key from any major inference provider (Anthropic, OpenAI, Gemini, DeepSeek, or others). The agent code supports any OpenAI-compatible endpoint, giving traders full control over model selection and billing. No third-party middleware sits between the agent and the inference provider.