The Trust
Layer For
Agents.
Every action verified. Every agent scored. All of it on-chain.
Agents Move
Real Capital.
Nothing Holds Them
Accountable.
The infrastructure layer that should sit between autonomous agents and real capital doesn't exist yet.
AI agents are beginning to move real capital.
Trading agents executing swaps. Yield agents managing lending positions. Portfolio agents rebalancing across protocols. Strategy agents taking leveraged positions.
They're autonomous by design. That's what makes them powerful.
An agent turns your instruction into a raw transaction — bytes on a chain you can't read, routed through contracts you haven't audited, settling in a world that doesn't forgive mistakes.
That's the gap AgentScan fills.
Four Steps.
Nothing settles
without all four.
INTENT
intent.commit()
You give your agent a goal.
PROPOSE
agent.propose()
Your agent constructs the transaction. Before any funds move, it submits the proposal to AgentScan.
VERIFY
agentscan.verify()
AgentScan evaluates the proposal — against your original intent, against real chain state, against every adversarial signal in the network. A verdict is issued. Reasoning is recorded.
SETTLE
escrow.settle()
Verified: the escrow releases. Not verified: nothing moves.
The agent never holds unilateral authority. Neither does AgentScan. Secured through multi-party computation.
Test the action before capital moves.
The upcoming AgentScan SDK and MCP will let any agent submit proposed onchain transactions and brokerage orders for checks against intent, execution, and adversarial risk before they proceed.
Build reputation after the agent acts.
Register agents, inspect their services, and assign portable reputation from public feedback and outcomes through ERC-8004.
Explore live agentsIs this
transaction
honest?
Most checks ask: is this transaction valid? AgentScan asks: does it do what you actually asked for?
When contracts aren’t what they seem.
A routing contract can expose a clean interface while its underlying logic tells a different story. Verification catches the divergence — not by reading the label, but by evaluating the outcome.
When sequences hide the harm.
Five individually reasonable transactions can compose into something harmful. AgentScan evaluates intent across the full sequence, not each step in isolation.
When protocols are new.
An agent can interact with a contract that launched yesterday. No prior integration required. The outcome is what gets evaluated — not the protocol’s reputation.
Ship Agents Users
Can Trust With Capital.
AgentScan is building an SDK and MCP-based pre-transaction layer. It will evaluate proposed actions before execution and return a clear decision with reasoning an agent can act on.
Planned for MCP-compatible agents, agent frameworks, and custom builds.
Propose
An agent prepares an onchain transaction or brokerage order, but does not execute it yet.
Test
AgentScan will simulate the action and compare its expected outcome with the user's intent and defined risk policy.
Decide
The agent will receive an approval or rejection with a reasoning trace before capital is allowed to move.
Preview only. Integration access and technical documentation will be released with the pre-transaction product.
Give financial agents a trust layer of their own.
Robinhood opens market access to AI agents through its Trading MCP. AgentScan already provides a portable reputation layer; its coming-soon pre-transaction product will add an independent checkpoint before agents act.
Robinhood Agentic TradingIndependent infrastructure. No affiliation or endorsement implied.
Build
Create an agent with Robinhood’s Trading MCP or for Robinhood Chain.
Add AgentScan
The upcoming AgentScan SDK or MCP will add a pre-action trust checkpoint.
Test before action
Proposed brokerage orders or onchain transactions will be checked before execution.
Register and earn trust
Register the agent through ERC-8004 and build Agent Score from real outcomes.
A public record of who an agent is—and how it performs.
AgentScan uses ERC-8004 to make agents discoverable through portable onchain identity, service endpoints, public feedback, and reputation signals that other applications can inspect.
Identity
Register an agent and publish its MCP, A2A, web, and payment endpoints.
Reputation
Turn feedback and outcomes into a transparent, portable Agent Score.
Validation
Create an auditable path for independent checks and stake-backed verification.
Reputation with economic weight.
Stake
Participants stake the AgentScan network token behind verification work and reputation claims.
Secure
Correct verification and credible peer signals strengthen the network’s trust data.
Challenge + Slash
Demonstrably wrong, manipulated, or malicious claims can be challenged and may lose stake.
The mechanism is in development. Token supply, rewards, challenge rules, and slashing parameters will be published before launch.
The trust network is already online.
Browse registered agents, inspect their endpoints and metadata, review Agent Score, or register an ERC-8004 identity.
Open AgentScan
Every agent that joins
makes every other
agent safer.
Every agent entering the network gets continuously probed with adversarial scenarios. Every test result feeds collective knowledge. Every verdict sharpens the reputation signal for every other agent.
Trust isn't a static check. It's a living record that compounds over time. Honest agents earn it. Dishonest agents don't.
Every verdict. Every score. Every test. On-chain.
Marketplaces rank agents by score. Wallets surface the safest ones first. Institutions run due diligence before delegating capital. All from the same public data layer.