What Relayline is, and what it is not
Relayline is a settlement layer that refuses to complete a transaction an agent did not declare in advance. It sits between an autonomous agent and Solana, and it produces a record pairing what the agent said it would do with what actually happened.
What problem it came from
Teams running autonomous agents onchain all reach the same wall. A transaction hash proves a transfer occurred. It does not carry the reason, the constraint the agent believed it was operating under, or whether the outcome matched the intent. When something goes wrong, the only evidence is the thing that went wrong.
The usual answer is to shrink the agent: lower limits, more approvals, a person watching a dashboard. That works, and it also removes the autonomy that made the agent worth building.
How it works, without the cryptography
Before an agent moves funds, it signs a statement of intent: the action, the counterparty, a value ceiling, and a reason. Relayline compiles that statement into an onchain constraint and submits the transaction inside it.
If execution stays within the declared bounds, the declaration and the result are written together. If it does not, the transaction reverts rather than partially filling. The record is readable by anyone holding the transaction signature.
What it deliberately does not do
Relayline does not judge whether an agent should have wanted something. A declaration that is honest and harmful will settle exactly as declared. It constrains the gap between stated intent and executed outcome, and nothing else.
It is also not a monitoring product. There is no dashboard that needs watching, because enforcement happens before settlement rather than after.
How to verify any of this
The constraint program is deployed and immutable, and its address is published on the tokenomics page. The specification is public. Any settled transaction can be resolved to its original declaration using only its signature, without asking Relayline for anything.
The team is anonymous. That is a fact worth weighing rather than one worth dressing up: the argument for the system rests on the published program and the spec, not on who wrote them.
Relayline is a Solana execution layer that requires autonomous agents to declare their intent before moving funds, recording the declaration and the result together.
Back to Relayline