Prepare Your Team

When more than one person runs Verdictan, decide who owns what before traffic depends on it. This guide takes a team from one working gateway to shared use: owners for each job, members with tested access, and one request that everyone can find.

Start when one person has finished the quickstart. You also need an organization administrator and the list of people to invite.

1. Assign an owner to each task

Write down an owner and a backup for each job. In a small team, one person can own several.

Job	Responsibility
Configuration	Approves changes, watches rollouts, and decides when to roll back
Request review	Investigates unexpected blocks and handles review items in Inbox
Evidence	Collects the records for an incident or a review
Spending	Reviews Usage and sets budgets and token limits
Access	Invites members, assigns roles, and removes access when people leave

2. Invite people and test their access

1. In IAM → Members, invite each person.
2. Create teams for people who share a job.
3. Assign the roles they need.
4. Ask a member who is not an owner to sign in and do one task their role allows.
5. Check that the same member cannot do an administrative task their role does not allow.

A sent invitation is not a finished setup. The member must accept it and confirm their access in the right organization.

3. Verify a shared request

1. Write down the test agent, gateway, organization, and region.
2. Give the gateway its provider key through its environment or an organization secret. A secret also needs agent access.
3. Give the test application its own application token.
4. Send the request from the quickstart.
5. With capture on, find the request in History and check its decision and usage.

A reviewer or administrator can also find the decision event from a CLI session in the same region:

verdictan events tail --since 10m --json

Match it by request ID. The provider key, gateway runtime token, and application token each do a different job, so keep them separate.

4. Agree how to handle problems

Choose a review rhythm that fits your traffic, and write down:

• Who handles an unexpected block or escalation, and how fast.
• Who can approve a configuration change or a rollback.
• Which request content you capture, and for how long.
• Who reviews spending and hears about budget problems.
• How access changes when someone changes role or leaves.

Link your team notes to request investigation, escalation review, and configuration changes.

Ready for shared use

You are ready when, in the right organization and region:

• A member who is not an owner can do their assigned work.
• A model request returns a response.
• Every job has an owner.

If something fails

Problem	What to check
A member signs in but cannot open a page they need	Organization membership, team membership, and role permissions
An escalation cannot be claimed	The member's escalations:claim permission and the owner of the item
A model request fails	The error, the application token, the gateway connection, and the provider credential
A request succeeds but is not in History	Capture settings, the request family, retention, and the selected region

Next steps

• Task Guides: the next workflow for your team.
• Usage Costs and Budgets: spending and its controls.
• SSO and SCIM: company sign-in and provisioning.