Read memory and playbooks
Inspect real decisions and understand the current Memory Market boundary.
A strategy's memory helps explain what it read, why it called a side, whether the gate allowed an order and what happened afterward. Its playbook is the creator's optional public description.
These are useful evidence to inspect before copying. They are not interchangeable with settled trading returns.
Read a decision
- Open Strategies and select the strategy or its memory panel.
- Check the decision's Window and time.
- Read the model's side, confidence and reason where an AI decision is available.
- Check whether the execution gate held or sent it.
- Follow the actual fill and later settlement outcome when recorded.
A hold can be an intentional result, such as insufficient confidence or a limit being reached. A sent order means an attempt was submitted; the fill record tells you whether a position was obtained.
Read a public playbook
Open the playbook from a strategy card or copy drawer. The creator can describe their approach during publication. That text is public plaintext. Treat it as the creator's explanation and compare it with the observed decisions and results.
Publishing a playbook does not make its claims independently verified. The code and contract limits decide what the runner can execute, and the records show what it actually did.
What Memory Market supports today
The interface includes a Memory Market area, but a paid, encrypted market for sealed memories is not implemented. There is no complete buy, decrypt, sell or revenue-collection workflow to follow in this build.
The current usable pieces are public playbooks and actual stored runner decisions when their data service is configured. Do not send funds expecting the disclosure panel to deliver a private memory package.
Why memory can be empty
A newly published strategy may not have been read by a runner yet. A model can be unconfigured. The decision store can be unavailable. A Dry read deliberately sends no order and stores no decision, so it does not populate this history.
Read the shown state rather than assuming an empty panel proves inactivity or a successful track record. Agent architecture explains the distinction between the chain's strategy record, the live runner and its stored decision history.
Source notes
This guide follows the application code reviewed on 2026-09-07. Links point to that reviewed commit and require repository access. GitHub may show 404 if you are signed out or do not have access.