VaultMind's AI agent holds its own key and spends from this vault — but only inside the policy the owner set on-chain. Over the cap, the chain itself refuses.
An AI copilot manages the money, an on-chain vault enforces the rules. Three steps, no trust required.
Every vault holds its owner’s own funds. VaultMind does not pool funds and does not accept deposits from third parties. The website functions as an interface for configuration and monitoring; transfers occur on-chain between the vault and the recipient.
The vault owner sets the policy: a per-payment cap, a daily cap, a recipient allowlist, and a freeze switch. This public demonstration deployment is configured with tight limits (0.01 BOT per payment, 0.025 BOT per day) because it holds real mainnet funds with an autonomous agent attached. Production deployments are configured by their respective owners, including their own limits.
The agent can pay only within the policy. A payment that exceeds the caps, is directed to a recipient outside the allowlist, or is attempted while the vault is frozen, reverts on-chain.
Limits shown here are the policy settings of this deployment.
This dashboard presents a demonstration deployment operated by its owner. VaultMind does not provide custody services and does not accept deposits from third parties. Each vault deployment is funded and operated solely by its respective owner.
These limits are the vault owner’s own policy settings, read live from the contract. This public demo runs deliberately tight limits because it holds real mainnet funds with an autonomous agent attached. Production deployments are configured by their respective owners, including their own limits.
No wallet connected. Connect a wallet to run the core VaultMind flow from this page: the owner can fund the vault, manage the recipient allowlist and pull the freeze switch, and the registered agent can pay inside the policy caps. The live reads above keep working without a wallet.