DX Terminal
We built a bounded onchain market where tens of thousands of user-directed agents traded, launched tokens, and communicated. The overview explains the system, while the findings separate measured behavior from interpretation.
Research
We publish deployment records, engineering methods, evaluation protocols, and reusable data for people building and assessing AI trading agents.
Follow new research via RSSResearch programs and deployments
These projects connect first-party deployments to the harness, runtime, controls, and evals we use around the model.
DX Terminal Pro
We ran a 21-day real-capital deployment on Base and preserved the path from user instruction through execution and settlement. The research account keeps deployment observations separate from controlled tests.
Trading-agent harness and runtime
Our architecture begins with an authenticated mandate and carries typed actions through policy validation, execution, settlement, reconciliation, and trace-based evaluation.
Trading-agent evaluation registry
We publish benchmark cards, harness-transfer tests, state and memory fixtures, and versioned data so readers can inspect the method and evidence class behind each result.
Public datasets
Choose published measurements for observed findings, source reviews for inspected papers and code, or evaluation methods for tests you can reuse. Each entry links its evidence and download formats.
Dataset
Formats
DXRG Open Trading-Agent Source Review
A dated comparison of two open trading-agent repositories, with pinned code references and explicit unverified questions. Source review only; no framework was executed or performance validated.
Source reviewVersion 1.0.0Published 2026-09-07Updated 2026-09-07
DX Terminal Pro Market Behavior Observations
Four aggregate cascade and two-sided-flow measurements from one bounded 21-day historical live deployment, with frozen definitions, source locators, privacy exclusions, and evidence limits.
Published measurementsVersion 1.0.0Published 2026-08-30
Presented inDX Terminal Pro
DXRG Trading-Agent Benchmark Audit Registry
Eight review units applied to three registered trading-agent benchmarks, with three frozen primary-source audits and exact disclosure locators.
Source reviewVersion 1.4.0Published 2026-08-12Updated 2026-08-24
DXRG Trading-Agent State and Memory Evaluation Fixtures
Eight unrun fixtures for measuring state transitions, memory supersession, policy response, and reconciliation boundaries.
Evaluation methodVersion 1.0.0Published 2026-08-03
DXRG Trading-Agent Prompt-Compilation Ablation Registry
Eight unrun comparison templates for testing prompt-compilation changes while freezing the model, state, memory, action schema, policy, execution adapter, and evidence class.
Evaluation methodVersion 1.0.0Published 2026-07-31
DXRG Trading-Agent Harness-Transfer Evaluation Card
Eight controlled comparison templates for separating model, prompt compilation, state, memory, action schema, policy, execution, and environment effects in trading-agent evaluation.
Evaluation methodVersion 1.0.0Published 2026-07-28
Presented inHarness-Transfer Tests for AI Trading Agents
DXRG Trading-Agent Trace-Derived Regression Registry
Eight versioned cases for turning linked trading-agent traces into bounded regression tests with frozen components, named interventions, diagnostics, downstream checks, and evidence-class limits.
Evaluation methodVersion 1.0.0Published 2026-07-24
Presented inUsing Trading Traces to Improve Agent Behavior
DXRG Trading-Agent Execution and Reconciliation Test Matrix
Eight failure fixtures for final-payload binding, request identity, ambiguous timeouts, acknowledgements, partial fills, fees, reconciliation, and controlled recovery.
Evaluation methodVersion 1.0.0Published 2026-07-24
Presented inFrom a Typed Trading Action to Settlement
DXRG Trading-Agent State and Memory Test Checklist
Eight failure fixtures for evaluating state identity, freshness, portfolio reconciliation, order lifecycle, venue state, memory provenance, action binding, and post-settlement feedback.
Evaluation methodVersion 1.0.0Published 2026-07-22
DXRG Trading-Agent Mandate Compilation Evidence Table
Each record states the evaluation design, intervention or setting, fixed components, observed measure, interpretation, and evidence boundary. Controlled pre-launch interventions and historical live setting gradients remain separate evidence classes.
Published measurementsVersion 1.0.0Published 2026-07-21
DXRG Agentic Trading Guardrail Matrix
A design and evaluation matrix for the deterministic controls inside a trading-agent harness, including mandate compilation, correct state, typed actions, validation, execution, and trace-level feedback.
Evaluation methodVersion 1.1.0Published 2026-07-19Updated 2026-07-20
DXRG Agentic Trading Benchmark Card
A disclosure and evaluation template for autonomous and semi-autonomous market systems. Completing the card does not establish profitability, safety, or suitability.
Evaluation methodVersion 1.1.0Published 2026-07-18Updated 2026-07-20