Participating in alpha is an opportunity to give precise feedback

Participate in alpha to test specific workflows and report precise observations. Access, referral participation, or an attractive example does not establish that a strategy has a verified edge. We would give feedback that helps improve the product while keeping our performance claims attached to measured outcomes.

DXAP's current FAQ describes referral-code access and live Hyperliquid trading. Its releases distinguish deployed entries from Upcoming changes. This is the public alpha scope inspected October 3, 2026, with an evolving interface and capability set.

A useful report about a strategy proposal

Illustrative report: an owner asks for a tighter market scope in chat. A proposal appears, the owner confirms it, and the saved settings show the intended symbols. A later turn attempts an instrument outside that scope and is rejected.

The report should contain the request, the reviewed change, the saved result, and the later action outcome. That sequence helps distinguish a proposal-display problem from a settings-save problem or a model proposing an inadmissible action. The rejected action may show a policy check working while raising a separate question about model behavior.

Keep private account identifiers and financial details in the appropriate private support context. A public discussion can describe the mechanism and its evidence class without exposing another user's activity.

State intended and observed behavior

Use a short structure: what triggered the issue, what you expected under the documented workflow, what you observed, and which records support it. Include the date and relevant public release version when known. If an error appears only once, say so. If several recorded cases show the same issue, state the denominator.

For strategy feedback, identify the instruction version and required evidence. “The agent made a bad trade” merges prediction, policy, and execution. “The turn cited an expired event as current evidence” gives a team a narrower failure to investigate.

Our trace feedback article explains stage-specific diagnosis. Our evaluation guide explains why a performance claim needs population, period, costs, and an appropriate comparison.

Share what your evidence supports

A workflow observation can support “I could review and approve the change.” A venue-confirmed fill can support “this order executed at this price.” A positive account interval can support a bounded description of that interval when cash flows and costs are reconciled. None of those alone establishes general profitability.

Keep referral incentives separate from the reason you recommend a workflow. If you publicly discuss participation, explain your actual relationship and the evidence you observed. A referral mechanism measures acquisition and fees, while a matched evaluation measures a different question.

The most useful alpha contribution is often an inconvenient counterexample: a missing measurement, a confusing approval state, or an instruction conflict that can be reproduced. Those reports help a product earn stronger claims through later verification. Participation is the beginning of that evidence collection, rather than its conclusion.