Dispute-ready records for AI-agent commerce.

Verdra captures who authorized an agent, what it was allowed to do, what it actually did, and what counts as evidence if the transaction is challenged.

Open the capture simulator → Open the Record console
Live · Verdra Record Evidence Package
EP-2026-0424-A23F
acme-procure-7 → vendr-fulfill-2 · $1,847.20 USDC
Capture pipeline 0 / 4 sealed
01
Transaction
The payment itself.
who paid whom · amount · rail · reference
T+0 ms
02
Configuration
How the agent was set up at dispatch.
prompt hash · tools · model · spend cap
T+4 ms
03
Behavior
Why the agent acted the way it did.
decision trace · tool calls · latency
T+12 ms
04
Capability
What the agent was authorized to do.
signed delegation · scope · expiry
T+18 ms
Sealed in T+0 ms BLAKE3 · stored in your bucket · countersigned
In plain English

AI agents are starting to spend money on behalf of companies — paying invoices, settling trades, ordering inventory. When something goes wrong, the company has to prove what the agent was told to do, what it actually did, and who authorized it.

Verdra captures that proof at the moment the agent acts, signs it with the company's own keys, and stores it in the company's own systems. Think of it as a flight recorder for an AI agent's transactions — there if you ever need it, invisible if you don't.

2023 → 2026 · agent infrastructure shipped
MCPagent ↔ tools A2Aagent ↔ agent x402on-chain settlement AP2payment mandates TAPagent trust score VIscheme dispute hooks LangSmithframework observability LangfuseOSS agent tracing

Verdra is the dispute-ready record.

"Agentic commerce needs legal infrastructure, and the courts are coming." Bridget McCormack (CEO, AAA) & David Hoffman (Penn), April 22 2026
Verdra Record

DR-grade evidence, captured at dispatch.

Verdra Record captures the dispute-ready record at dispatch every time an agent transacts. Four strata, signed by the agent's own keys and sealed: transaction, configuration, behavior, and capability.

01Transaction. Protocol-level call: who, when, payload, hash.
02Configuration. Prompt, tools, model, version at dispatch.
03Behavior. Counterfactual choices the agent would have made.
04Capability. Platform tools and routes the agent could invoke.

SDK: Python + TypeScript · Signed: BLAKE3 + RFC 3161 · Storage: enterprise-owned

Works with: LangGraph · LangChain · AutoGen · CrewAI · Semantic Kernel · OpenAI · Anthropic · Google · custom runtimes

app.verdra.com/record
Open the Record console → Talk to engineering →

If you run agents that transact, you're already in the dispute-resolution (DR) risk perimeter.

Verdra Record is in pilot for design-partner deployment to be DR-ready ahead of the EU AI Act's August 2, 2026 enforcement date. Pricing is engagement-based — talk to us for a pilot quote.

Start a pilot → Open the Record console