Agents your developers wrote
Core SDKs for TypeScript, Python, .NET, Java, Rust, and Go wrap the action your application owns. The released catalog includes 35 Enforce adapters for documented framework boundaries.
Integrations
Keep how your agents are built. Add the rules they answer to.
Some agents are written by your developers, some are assembled in Copilot Studio, and some pay for things on their own. Each one connects to the same business rules, sends its exceptions to the person who can approve them, and leaves one record that risk, compliance, and audit can all read.
Start from how the agent was built
Agents built in Microsoft Copilot Studio
Business teams keep building in Copilot Studio. Before a connected action changes a loan, an account, or an order, it gets checked against your rules.
Agents that work over MCP or in a browser
Agents can look up records, draft a rule, or check an action before they take it, with the same access the signed-in person already has.
Agents without a VeriProof SDK can send records to the API over HTTPS. Operators who prefer a terminal can use the VeriProof CLI for setup, lookups, and evidence exports.
Let your agent pay, within limits you set
An agent that buys data, pays suppliers, or checks out for a customer needs a spending rule and someone to call when a payment falls outside it. VeriProof checks the payment before your wallet or provider moves any money, and sends exceptions to the budget owner.
Paid API calls over x402
The agent pays per request for data and tools. Rules cap what it spends and which networks it can use.
pay.sh and Machine Payments Protocol
The agent picks a paid provider from a catalog. Rules check that the provider is verified and that the session stays under its cap.
AP2 mandates
The agent checks out on someone's behalf. Rules hold it to the payee, amount, and expiry in the mandate it was given.
Stablecoin supplier payments
The agent pays suppliers in stablecoins. New counterparties and large cross-border transfers go to treasury first.
Work alongside the tools your teams already use
Microsoft Agent 365
Agent 365 keeps your agent inventory, identities, and security controls. VeriProof decides whether each business action is allowed.
Runtime guardrails
Keep the guardrail product you run today. Send its result with the policy check, and blocked cases reach a person who can decide.
Langfuse
Engineers keep debugging in Langfuse. The same traces show the policy result for each checked action.
Splunk
Security analysts search policy outcomes in Splunk and follow a link to the VeriProof record behind an alert.
GRC and audit
Compliance and auditors work in their own VeriProof portals, and evidence packs travel to the systems they already use.
Your model provider
VeriProof's own AI features run on Azure OpenAI or an OpenAI-compatible endpoint you choose, so your provider decision stays with your team.
Coming soon: Splunk Observability Cloud, Vanta, and OpenTelemetry metrics export.
Reach reviewers where they already work
An agent that pauses for a person is only as quick as that person's answer. VeriProof sends each review to the VeriProof Inbox and by email, and to Slack or Microsoft Teams if your team lives there. Each Slack or Teams message updates as the review moves, so nobody acts on one that's already settled.
- Slack. Reviewers decide from their own message and can provide typed answers within a data limit you set. Decisions are recorded in VeriProof. See the Slack integration →
- Microsoft Teams. Reviewers decide from their own Teams card, including typed answers within a data limit you set, and every decision is recorded in VeriProof. See the Teams integration →
- Your own systems. Subscribe to signed review events and send notices through the tools you already run.
Same record, six portals
Every connected agent feeds the same handoffs
However an agent connects, its decisions land in portals built for the business owner, risk and compliance, governance engineers, developers, auditors, and administrators. Work moves between them through built-in handoffs.
An agent action needs a person
Reviewer with authority
A rule needs data the agent does not send
Developer
A governance engineer changes a rule for a production agent
Administrator or Compliance Officer
A control is not met
Assigned owner
A risk needs a business decision
Business Owner
An audit begins
Auditor
Need a connection we don't list?
Tell us which system your agent runs in and what it needs to do. We'll tell you which connection fits and reply within one business day. If you build or run agents for clients, the partner program covers implementation and advisory work.