Assurify is the on-chain coverage layer for AI agents operating capital on Solana. Oracle-verified payouts in USDC for financial loss, per-transaction slippage, and data breach events — settled in under a minute, without adjusters, brokers, or paperwork.
Parametric triggers underwritten on-chain. Pricing is risk-based, computed from agent autonomy, capital under management, and historical on-chain behavior.
Coverage activates when an agent's wallet balance breaches the pre-set drawdown threshold. Switchboard monitors on 60-second intervals; Pyth resolves USD-denominated loss; USDC settles automatically.
Coverage activates on every transaction. Pyth records the asset price at submission and at execution — if the adverse delta exceeds the configured threshold, USDC settles within the same slot. Built for HFT and market-making agents.
Coverage for agents that leak, transmit, or mishandle protected data. Policyholders submit cryptographically signed execution logs as evidence. A staked reviewer panel cross-references the on-chain log commitment and settles within 72 hours.
No intermediaries. No manual reviews for parametric triggers. No counterparty risk. The same Solana infrastructure that settles DeFi trades settles your coverage.
Assurify is an Anchor program deployed on Solana, with dual-oracle verification, a USDC-denominated reserve pool, and a staked review panel for hybrid-settled policies.
Pyth Network provides sub-second price feeds for valuation. Switchboard provides on-chain wallet monitoring at 60-second intervals. The combined stack means every parametric trigger is independently verified by two battle-tested oracle networks before payout.
The protocol maintains a reserve backing of 1.5× outstanding coverage, funded by external liquidity providers earning 8–14% APY. Idle capital is deployed to Kamino Finance for additional base yield, with a seven-day unbonding period.
A parametric reinsurance tranche through Uno Re provides catastrophic-loss coverage above protocol reserves. This institutional-grade backstop activates only under systemic drawdown conditions, protecting both LPs and policyholders.
Data breach claims are settled by a panel of five independently staked reviewers. Approvals require 3-of-5 consensus. Reviewers voting in the minority on clear cases are slashed, aligning incentives with protocol integrity rather than individual claim outcomes.
// Parametric financial loss trigger pub fn settle_claim( ctx: Context<SettleClaim>, ) -> Result<()> { let policy = &ctx.accounts.policy; let oracle = &ctx.accounts.pyth_feed; // Read verified Pyth price let price = get_pyth_price(oracle)?; let balance = read_wallet( &policy.agent_wallet )?; // Compute drawdown vs threshold let loss = policy.baseline .checked_sub(balance * price) .ok_or(ErrorCode::Underflow)?; require!( loss >= policy.threshold, ErrorCode::NoTrigger ); // Transfer USDC payout to agent token::transfer( ctx.accounts() .into_transfer_ctx() .with_signer(&[b"reserve"]), loss.min(policy.limit), )?; emit!(ClaimSettled { policy: policy.key(), amount: loss, slot: Clock::get()?.slot, }); Ok(()) }
Every policy, premium, and payout is observable on-chain. The following metrics are drawn from the live protocol state as of the most recent block.
Premiums are computed on-chain from each agent's risk profile — autonomy level, capital under management, and historical operational behavior. No brokers, no hidden fees, no manual underwriting cycles.
Common questions about the protocol, coverage triggers, payout mechanics, and LP participation. For anything not covered here, reach out directly.
hello@assurify.io →Connect Phantom and have coverage live in under two minutes. No paperwork, no brokers, no waiting room — just parametric protection that settles when your agent needs it.