Agenor Attest · self-serve scoping

Scope your agent trust surface

A few questions about your agents. It computes exactly what Attest would run, at the access level you choose, and generates a spec you can email me or bring to a call. This is the same interview I'd run on a call, minus the call.

🔒 Runs entirely in your browser. Nothing is sent or stored, no sign-up.

1. What do your agents do, and what stage are you at?

One line. e.g. "Agentic marketing OS that runs ad campaigns, soft launch."

Pre-launch Soft launch GA / in production

2. What can your agents actually DO? (not just say)

Check everything that applies. This sets your risk tier.

3. What runs your agents?

Not sure? Pick "Unsure", it's a one-line answer from your engineer and we'll flag it.

4. Where do agent runs get logged today?

5. Connectors, tenancy & approvals

Roughly how many external services do agents connect to, and how is it run?

A human approves high-risk actions There are budget / quota caps Multi-tenant (many customers, shared stack) Agents delegate to other agents

6. How much access are you comfortable giving Attest? ← this picks the run mode

We always start at the least access that produces value, and step up only with your explicit OK.

Tier 1, StaticWe read your declared config only. Zero runtime access. Tier 2, Observe (recommended)A read-only copy of your traces. Nothing in your request path. Tier 3, Replay on stagingWe probe a staging copy. Never touches production. Tier 4, Live read-only offenseNon-intrusive probes against prod. Never mutates anything.

7. Who are you? (so we can send the spec back)

Scope your agent trust surface

Your Agenor Attest Onboarding Spec