lyudik.rofd.io
Contact me
Language:EN中文

Growth · 2 min read

️ "2 of 3 Signatures Required" — Multi-sig UX Nightmare

✍️ "2 of 3 Signatures Required" — Multi-sig UX Nightmare
Gnosis Safe is the gold standard for treasury security. Its default UX assumes every signer has a CS degree and infinite patience.
The coordination failure mode is predictable: transaction proposed, signer 1 signs, signer 2 doesn't know about it, transaction ages in queue, eventually expires or becomes outdated, cycle repeats. The Safe team's own analytics showed average signing time of 48 hours for 2-of-3 multisigs before their dashboard redesign.
The underlying problem is context collapse. Multi-sig interfaces show pending transactions as technical artifacts (nonce, data, value) rather than business actions ("Pay $50K to vendor X per contract Y"). Signers don't know why they're signing, what depends on this transaction, or who else needs to act.
After implementing their improved signer dashboard—which surfaces transaction intent, shows signing progress with names/avatars, and integrates notification channels—Safe reported average signing time dropped to 4 hours. That's a 12x improvement from UX changes alone.
Production requirements for human-usable multi-sig:
→ Transaction intent display, not just parameters. Integration with transaction decoders to show "Swap 100 ETH for USDC on Uniswap" rather than raw calldata.
→ Signing progress visualization. "Alice signed 2 hours ago. Bob: pending. Charlie: pending. Need 1 more signature to execute." This pattern is documented in the Safe SDK examples.
→ Integrated communication. Click to notify pending signers via email, Telegram, or Slack. The transaction context travels with the notification.
→ Deadline awareness. If the transaction involves time-sensitive operations (auctions, governance deadlines), surface urgency prominently.
The multi-sig security model is sound. The coordination model breaks down at human scale without proper tooling.
→ MayWap | Designing multi-sig that humans can use