Runtime DeFi Guard

Reviews Safe transactions before signing and captures proposal context for Runtime DeFi Guard reports.

Runtime DeFi Guard reviews a Safe transaction at the moment that matters: after it is proposed, and before your co-signers approve it. It reads the proposal directly on app.safe.global, reconstructs exactly what your wallet is being asked to sign, and turns it into a report your whole team can read in seconds. What it captures, automatically: The full Safe transaction as proposed: destination, value, calldata, nonce, operation. The dApp that requested it, confirmed through Safe's own WalletConnect origin data. The typed data your wallet will actually sign, with the signing hash recomputed locally and checked against Safe's records. Why signers keep it on: A transaction is easy to approve and impossible to undo. Guard gives you an independent second read inside that window. It surfaces where a request really came from, so a familiar-looking transaction from an unexpected source does not slip through. It flags when what the interface shows differs from what your wallet is signing. Built for how funds and DeFi teams operate: Works alongside Safe. Safe stays the source of truth; Guard never replaces it. Never touches your keys. It cannot sign, and it never records or transmits your wallet signature. No custody, no execution authority. It reads, reviews, and reports. GREEN is a reviewed transaction, not a guarantee. Guard is an independent signal that makes the last step before signing legible, so approval becomes a decision rather than a reflex. Runtime DeFi Guard is part of Runtime Research, the technical intelligence layer behind Runtime Capital. More at https://www.runtime-guard.io.