FAQ#
The Basics#
What is Hightop?#
Hightop is digital banking for AI agents, made simple for humans. You fund an account, give each agent its own rules — spending limits, approved recipients, asset restrictions, timing controls — and let your agents pay, get paid, earn, borrow, and convert through the API. You use the app. Agents use the API. Both connect to the same system underneath, and key rules are enforced onchain by open-source smart contracts. See How Hightop Works.
Who is Hightop for?#
Anyone running AI agents that need to handle money — paying for compute, data, or vendor services, collecting revenue for work agents perform, managing treasury and yield, or borrowing working capital. If your agents need financial autonomy within boundaries you control, Hightop is built for that.
Do I need to know anything about crypto or blockchain to use Hightop?#
No. Hightop handles the wallet infrastructure, stablecoin transactions, and gas fees behind the scenes. You interact with a consumer-grade app and familiar money rails like bank transfers and ACH. The blockchain is where key rules are enforced, but you do not need to understand or interact with it directly. See Wallet Model for why Hightop chose this infrastructure.
What does "digital banking for AI agents" actually mean?#
It means agents get a real money layer — the ability to pay, get paid, earn yield, borrow working capital, and convert between supported assets — inside boundaries you define and the system enforces. Without something like Hightop, agents either have no access to money, or they have unrestricted access, which is dangerous. Hightop solves that tradeoff: agents can operate autonomously, but only within lanes you define and the system enforces.
Trust and Safety#
How is my money protected?#
Your funds live in a smart contract wallet on a blockchain — not on Hightop's balance sheet. The wallet enforces your rules onchain: agent permissions, spending limits, approved recipients, asset restrictions, and timing controls are all checked by the smart contract before funds move. Your control wallet's keys are not stored on Hightop's servers — they live within Turnkey's infrastructure. High-risk changes like adding a new trusted destination are delayed, giving you time to cancel if something looks wrong. See Security and Control.
What does "onchain enforcement" actually mean?#
It means the rules that protect your money are enforced by open-source smart contracts running on a public blockchain — not just stored in Hightop's database. The smart contract checks every agent action against your configured rules before funds move. The code is open-source, the execution is public, and anyone can verify the rules independently. This is the difference between "we promise this limit is enforced" and "here is the code, and here is the proof that it ran." See Why Onchain Enforcement Matters.
Is Hightop like FTX or Celsius?#
No. Those platforms held customer funds on opaque balance sheets and made hidden bets with deposits. Hightop is built on transparent, onchain smart-contract infrastructure — your funds live in visible wallets and contracts, not on a company's balance sheet. Hightop does not rehypothecate your assets or make private bets with your money. If your funds earn yield, it is because you chose to deposit into Earn, and those funds go to visible, auditable protocols. DeFi protocols carry their own risks, and Hightop is not FDIC insured.
What happens if Hightop the company goes away?#
Your assets live in a smart contract wallet on the blockchain, not on Hightop's servers. The smart contracts are open-source and deployed onchain — they continue to exist and enforce rules independently of Hightop's infrastructure. The practical recovery path depends on your custody setup: if you have upgraded your control wallet to a hardware wallet or Safe, you can interact with the smart contracts directly through available interfaces. With the default Turnkey embedded wallet, your access path depends on Turnkey's infrastructure remaining available.
What if a smart contract has a bug?#
The smart contracts have been independently audited — see Audits for the full reports — and the code is open-source and published on GitHub for anyone to review. No software is guaranteed bug-free, but independent audits plus open-source code plus onchain transparency make issues discoverable and verifiable. The Hightop team also built the underlying protocols (Underscore Protocol and Ripe Protocol), so the team developing the product is the same team that built and maintains the contracts.
Is Hightop FDIC insured?#
No. Hightop is not a bank and is not FDIC insured. Your funds are protected by onchain smart contracts, not by government insurance. Reduced risk is not zero risk — DeFi protocols carry their own risks, including smart contract risk, market risk, and protocol-specific risks. See Security and Control.
What if my email is compromised?#
With the default Turnkey embedded wallet, your email-based access path is the route to your control wallet. If your email is compromised, an attacker could potentially take over the control wallet. Delayed changes like adding new trusted destinations and activation delays on new agents still create a response window, but that protection applies to new changes, not destinations that are already active. If you want to remove this risk, you can upgrade your control wallet to a hardware wallet or Safe, which eliminates the email-based access path entirely. See Security and Control.
Can Hightop staff access my funds or change my rules?#
No. Your control wallet's keys are not stored on Hightop's servers. Hightop cannot override the onchain rules, move your funds unilaterally, or change agent permissions on your behalf. The smart contract is the authority — it enforces the same rules for every request, regardless of who submits it. See Your Hightop Wallet.
Agents and Control#
What is a Hightop agent, exactly?#
A Hightop agent is not the AI itself. It is a named control profile — a financial lane you build for an external agent or workflow. It defines what that agent is allowed to do with money: permissions, spending limits, approved recipients, allowed assets, timing rules, and expiry. Your external agent (running in a framework like OpenClaw or Hermes, or your own code) talks to Hightop through the API, and the smart contract checks every request against that lane. See AI Agents.
Does my AI agent live inside Hightop?#
No. Your agent lives wherever you run it — your own infrastructure, an agent framework, or a cloud service. Hightop is where the financial lane is defined and enforced. Your agent talks to Hightop through the API. Hightop does not host, run, or manage your agent's logic.
What happens if my agent tries to do something outside its rules?#
The action is blocked before funds move. The smart contract checks every request against the agent's permissions, limits, approved recipients, allowed assets, timing rules, and payment-path rules. If any check fails, the entire transaction reverts — no partial execution, no "almost went through." See If an Agent Goes Off-Script.
What happens if my agent's API key is compromised?#
The attacker gets the same authority the agent had — and nothing more. They can use whatever permissions and limits you configured for that agent, but they cannot widen them, change the rules, add new recipients, or access other agents' lanes. Key boundaries are enforced by the smart contract, not by Hightop's servers. This is why narrow, purpose-built lanes matter: the tighter the lane, the less damage a compromise can cause. See Security and Control.
Can agents change their own rules or add new recipients?#
No. Only the control wallet can change rules, widen permissions, add agents, or approve new payment recipients. An agent cannot grant itself new permissions, create other agents, escalate its authority, or add trusted destinations. An agent can propose a new recurring vendor, but it cannot activate or confirm one — only the account owner can. This is enforced by the onchain control layer. See AI Agents.
How narrow should I make my agent's permissions?#
As narrow as the agent's actual job requires. A research agent that pays three vendors does not need access to Earn, borrowing, or trading. An inference agent with a $200 per-transaction budget does not need a $10,000 per-transaction cap. Start with the minimum permissions, lowest limits, and shortest expiry window that let the agent do its job — you can always widen later. See Agent Permissions and Limits.
Can I pause or remove an agent immediately?#
Yes. You can revoke an agent at any time through the app. Removal takes effect immediately — the agent can no longer act. There is no wind-down period where it retains partial authority. You can also let agents expire automatically by setting an expiry window when you create them. See AI Agents.
What is the difference between account-wide rules and agent-specific rules?#
Account-wide rules set the outer ceiling for every agent on the account. Agent-specific rules narrow that further for each individual agent. Rules only get tighter as they layer — an agent can never exceed the account-wide ceiling. For example, if the account allows $10,000 per day in total outflows, but a specific agent is capped at $1,000 per day, the agent is limited to $1,000. See Agent Permissions and Limits.
Payments#
What are the three payment paths and when do I use each one?#
Hightop has three outbound payment paths, each designed for a different trust level. Recurring Payments are for known vendors you pay repeatedly — you approve the vendor once, set limits, and let the agent handle the flow. One-Off Payments are for ad hoc payouts to new or one-time recipients — small payments can clear quickly, while larger ones sit behind a review delay. Trusted Transfers are for your own highest-trust destinations — fast transfers with no recipient-level caps or delays, but adding a new destination requires a days-long confirmation period. See Recurring Payments, One-Off Payments, and Trusted Transfers.
What payment methods does Hightop use?#
Hightop uses two main payment surfaces. Stablecoins, especially USDC, power programmable payments — recurring vendor payments, one-off payouts, trusted transfers, and machine-native flows like x402 and MPP. Hightop Cards power traditional merchant spend through virtual cards, while still drawing from the same Hightop account and control model.
What is the difference between a recurring payment and a one-off payment?#
A recurring payment goes to a pre-approved vendor with fixed rules — the relationship is set up in advance with per-vendor limits, allowed assets, cadence, and expiry. An agent can pay that vendor repeatedly within those boundaries. A one-off payment goes to any recipient, but larger amounts sit behind a review delay and can be cancelled before they clear. Recurring payments have lower friction for known vendors. One-off payments have more friction but allow ad hoc payouts. See Recurring Payments and One-Off Payments.
Why do one-off payments have a delay?#
Because they are more likely to go to unfamiliar destinations. The delay creates a review window — time for you to notice a bad destination, a rogue agent, or a compromised API key and cancel before funds move. Small payments below your configured instant threshold can still clear immediately. The delay is enforced by the onchain control layer, not just a front-end timer. See One-Off Payments.
What is a trusted transfer?#
The highest-trust payment path in Hightop. Trusted transfers are meant for destinations you trust deeply — your own wallets, reserve accounts, emergency locations, or connected bank accounts for withdrawal. Transfers to active trusted destinations can happen immediately, without the normal recurring-payment or one-off-payment recipient checks. Adding a new trusted destination requires a timelocked confirmation period. The list should stay short and deliberate. See Trusted Transfers.
Do trusted transfers bypass all limits?#
Trusted transfers bypass recipient-level friction once a destination is already active. If an agent initiates the transfer, that agent's own delegated limits still apply — for example, an agent with a $1,000 per-transaction cap cannot send $5,000 to a trusted destination. If the control wallet initiates the transfer, those agent-delegated caps do not apply, so any amount in the wallet can be sent immediately to an active trusted destination. The main protection is that adding a new trusted destination is delayed and controlled by the control wallet. See Trusted Transfers.
Can my agent receive payments?#
Yes. Agents can receive payments for work they perform — usage-based revenue, API charges, or service fees. That can happen through Hightop's normal stablecoin-native payment identity or through machine-native payment flows like x402 and MPP. See Emerging Payment Protocols.
Wallet and Account#
What is a control wallet?#
The control wallet is the only address that can change your account's rules — add an agent, update limits, approve a new payment recipient, or remove an agent. Think of it as the master key. No one else, including Hightop, can make policy changes without it. By default, Hightop creates your control wallet as an embedded wallet through Turnkey, tied to your email. You can later upgrade to another wallet (hardware, Safe) for stronger security. See Your Hightop Wallet.
What is the difference between my control wallet and my smart contract wallet?#
They serve different roles. Your control wallet is the address that sets and changes the rules — it is the master key. Your smart contract wallet is where your assets actually live — it holds your balances, enforces the rules onchain, and decides whether each agent action is allowed. The control wallet configures. The smart contract wallet enforces. Agents operate within the smart contract wallet's rules but cannot change them. See Your Hightop Wallet.
Do I need to manage private keys or gas fees?#
No. Hightop handles the wallet infrastructure and covers gas fees for normal product usage. You do not need seed phrases, private keys, or crypto tooling. The app handles everything behind the scenes. Under the hood, the rules still live onchain in open-source smart contracts — you get blockchain-enforced control without the crypto UX burden.
Can I upgrade my wallet security?#
Yes. You can move control of your account from the default Turnkey embedded wallet to a wallet you fully control — a hardware wallet like a Ledger or a Safe (a multi-signature wallet that requires multiple approvals to act). This removes the email-based access path entirely and puts control under your own keys or signing authority. Your assets stay where they are, your agents keep running, and nothing is disrupted. See Your Hightop Wallet.
What is Turnkey?#
Turnkey is a secure key-management platform. When you sign up for Hightop, it creates your control wallet as an embedded wallet through Turnkey, tied to your email-based authentication. The important part: your control wallet's keys live within Turnkey's infrastructure, not on Hightop's servers. That means a compromise of Hightop's servers alone is not enough to take over your control wallet. See Your Hightop Wallet.
Earning and Borrowing#
How does Earn work?#
Your funds go into a vault managed by Hightop-built AI agents that continuously optimize yield across approved protocols — routing based on risk, not just rate. You receive vault shares representing your proportional ownership. As the vault earns, your shares become worth more. The vault agents operate within the same onchain control model as everything else in Hightop — they cannot send payments, deposit into unapproved protocols, or exceed their boundaries. Withdraw anytime. See Earn Under the Hood.
What are the risks of using Earn?#
Earn deploys your funds into DeFi yield protocols. These protocols carry smart contract risk (bugs in the protocol code), market risk (changes in yield rates, liquidity, or conditions), and operational risk. Hightop's vault agents only deploy to pre-approved protocols — this restriction is enforced onchain — but approval does not eliminate protocol-level risk. Hightop Earn is not a savings account and is not FDIC insured. The performance fee is 20% on profits only, never on principal.
What is the difference between Core and Amplified vaults?#
Core Vaults are the default — passive, AI-managed yield with no borrowing or leverage. Amplified Vaults build on top of Core Vaults by borrowing stablecoins against your Core Vault shares and deploying the borrowed funds into additional yield. The borrowed funds always stay in dollar-denominated assets, which avoids the death-spiral risk of traditional leverage. Amplified Vaults target higher yield but carry additional risk from managed leverage and debt. See Earn Under the Hood.
How does borrowing work?#
You borrow against your existing assets, not your credit score. Your entire portfolio backs a single overcollateralized loan through Ripe Protocol — one position, one interest rate, one set of thresholds. Repay any amount at any time with no penalties. See Borrowing Under the Hood.
What is productive collateral?#
It means your collateral does not have to sit idle. Your Earn vault shares — the tokens you receive when you deposit into an Earn vault — can be used as collateral for borrowing on Ripe Protocol. Your assets earn AI-managed yield while simultaneously backing a loan. You are not choosing between earning and borrowing — you get both. This is also how Amplified Vaults work internally. See Borrowing Under the Hood.
What happens if my collateral drops in value?#
If your collateral drops too far in value, the protocol can reduce debt before full liquidation and, if needed, liquidate only what is necessary to restore the position to health rather than the entire position. You can also proactively add collateral or repay debt at any time to reduce risk. See Borrowing Under the Hood.
Practical#
How do I fund my Hightop account?#
Move money in through bank transfers, ACH, Apple Pay, or card-based onramps. Your funds land in one Hightop account, where they can immediately be used by agents, deposited into Earn, or held as collateral — no manual conversion or wallet setup required.
How do I withdraw?#
Withdrawals go to destinations you have activated in Account → Bank Accounts or Account → Crypto Addresses. Once a destination is active, transfers to it can happen immediately. Adding a new destination requires a timelocked confirmation period — the same delayed-activation model used by Trusted Transfers — so a new destination never becomes usable the moment it is added. For the step-by-step flow, see Fund and Withdraw.
What assets does Hightop support?#
Hightop supports stablecoins like USDC, established crypto assets like ETH and BTC (cbBTC), Earn vault shares, and other protocol-native tokens. Earn routes across approved yield protocols like Aave, Morpho, Euler, Moonwell, Fluid, and Compound. Convert flows are powered by onchain venues. For the fuller reference view, see Supported Assets and Venues.
What are x402 and MPP?#
x402 and MPP are machine-native payment standards that Hightop supports for agent-to-service and agent-to-agent commerce. x402, developed by Coinbase, builds around HTTP 402 Payment Required so APIs, apps, and agents can charge and pay for digital services directly over standard web requests. MPP, short for Machine Payments Protocol, is another standard from Stripe and Tempo for machine-native payments, including micropayments and recurring charges. In Hightop, they sit alongside the rest of the stablecoin payment stack rather than creating a second money system. See Emerging Payment Protocols.
What blockchain does Hightop run on?#
Hightop runs on Base, a Layer 2 blockchain built on Ethereum. Base offers low transaction costs (fractions of a cent), fast settlement, and the security guarantees of Ethereum underneath. Hightop users do not interact with Base directly — the app handles everything behind the scenes.
What are Underscore Protocol and Ripe Protocol?#
Underscore Protocol powers Hightop's wallet infrastructure, control layer, and earn vaults. Ripe Protocol powers the borrowing infrastructure. Both are open-source and built by the Hightop team. You do not need to interact with either protocol directly — Hightop wraps them in a product experience. But if you want protocol-level details, see Technical References for the full map of repos, technical docs, and live params tools.