Security on Vectr
Vectr handles wallets, API keys, trading and AI agents, so the rules are simple: you confirm an agent's transaction before anything signs, scheduled orders act only inside limits you authorise, keys are scoped, and every contract is on the explorer. The controls below are served live from the API's platform descriptor.
Custody
- Your own wallet
- Connect any EVM wallet and sign everything yourself — Vectr never sees a key.
- Vectr wallet (X login)
- Logging in with X creates a custodial wallet. Private keys are stored encrypted (AES-256-GCM).
- Key export
- Private-key export is disabled until it can be gated behind re-authentication. Need out? Contact support rather than waiting on a raw key over HTTP.
- Creator fees
- Curve fees accrue in the FeeCollector contract and are claimed by the token's fee recipient — the only balance that contract ever holds.
Transaction controls
- Confirmation first
- The agent turns your request into a typed intent and builds an unsigned transaction. You review and confirm it before your wallet signs and broadcasts it. The AI never writes raw calldata: it builds typed intents the backend validates on its own.
- Agent jobs over API, MCP and CLI
- Through the API, MCP server or CLI, an agent job returns the unsigned transaction and stops there: nothing is signed or broadcast until you submit it yourself.
- Scheduled orders
- Limit, stop, DCA and TWAP orders are authorised once, when you create them in the order form or give the agent their parameters. After that, eligible fills execute automatically from your wallet under the spending, recipient and contract controls you set.
- Selector + target whitelist
- Before signing, the wallet service checks the calldata selector against a per-action whitelist and the target against known platform contracts.
- Spending limits
- Optional daily and per-transaction USD limits, enforced server-side before anything is signed.
- Pause switch
- One switch halts all signing for the wallet.
- Recipient allowlist
- Restrict outgoing transfers to addresses you approve, with a cooldown before new recipients activate.
- Contract-call guard
- A switch to block arbitrary contract calls entirely, limiting the wallet to known Vectr actions.
- Anomaly detection
- Transfers well above a wallet's rolling average get flagged before they sign.
API keys
- Scoped keys
- Keys can be read only, read write, ip allowlist, recipient allowlist — and shown in full only once at creation.
- Read-only can't broadcast
- A read-only key is rejected at the signing layer, and an agent job run with it cannot create or cancel orders or turn on auto top-up: it can read data and prepare transactions, never submit or schedule one.
- Read-write keys can move funds
- A read-write API key is the automation path: it can submit transactions from your Vectr wallet without a person in the loop, inside the key's limits. A read-only key can never sign, submit or create orders. It broadcasts through /wallet/sign-and-submit, so treat it like a password and prefer read-only keys for data access.
- Instant revocation
- Revoke any key from your profile and it stops working immediately.
Contracts & source
- Not upgradeable
- Tokens and curves are EIP-1167 clones of fixed implementations — there is no proxy admin key that can rewrite them.
- Locked liquidity
- Graduation mints a full-range Uniswap V3 position into VectrLPLocker — it can never be withdrawn.
- Platform fees never pool
- The platform's 0.67% share is pushed to platform wallets in the same transaction; the FeeCollector only ever holds unclaimed creator fees.
- Fee cap on-chain
- No fee rate can be set above 5.00% — a bound enforced by the contract, bounding what even a compromised owner key could charge.
- Open source
- The contracts, CLI, MCP server and skills are public in the Vectr git.
Your controls, in the product
Spending limits, recipient allowlists, contract-call blocking and the pause switch live in your profile security tab. API keys are created and revoked there too. An agent's transaction waits for your confirmation in the Terminal; scheduled orders fill under these same controls.
A note on claims
This page describes what the code does — rendered from the platform descriptor on GET /config and the contracts on robin.etherscan.io. On-chain trading carries smart-contract, liquidity and automation risk — transactions are irreversible. If you find a vulnerability, report it through the source repo or the contact in /.well-known/security.txt. No third-party security audit has been performed; "verified" on this site means verified source on the explorer. The signing pipeline itself is explained stage by stage in how an AI agent trades on Vectr; the plain-language risks are in the risk disclosure.