Permission infrastructure for onchain agents
Autonomy,
with terms.
Your agent’s next move.
Your terms of execution.
We’re shaping a layer that connects what you authorize with what an agent can do onchain.
Explore the flow
Readable mandates.
Scoped actions.
Traceable receipts.
The next layer of delegation
More agency.
Clearer authority.
As agents start to act onchain, every instruction needs a boundary.
Which asset? Which action? How much? For how long? Writrail’s vision is to make those terms readable to people and enforceable at execution.
We’re starting with the connection between a user’s mandate, an agent’s action, and the record it leaves behind.
Follow the architectureOne continuous relationship
From permission
to proof of action.
Authority, made legible
Put intent into terms.
Define the agent, allowed action, spending cap and expiry in one mandate. The goal: know exactly what you’re handing over.
Take the idea for a run
Set the terms.
See what happens.
A small, interactive example of delegated authority. Change an action, run it, or revoke the mandate.
Ready when you are.
Run 20 USDC, then try 90. See how the remaining allowance changes the outcome.
Execution record
0 actionsYour first action will appear here.
Each record stays linked to this mandate.
The direction ahead
A common language
for delegated capital.
We want wallet teams and agent builders to share a clearer way to request authority, check its limits, and follow what happens next.
Wallet integrations & execution tooling
Coming soon