PRODUCT WALKTHROUGH · 2:07
See an investigation, team discussion, operator direction and request evidence in one run.
The Room brings engineers, security engineers and BattleTest AI into the same live application assessment. Before starting, define the permitted targets, choose the testing permissions and provide the access the assessment needs. This setup overview shows a target, its permission tier and its environment. Scope and public-target authorisation are managed in Settings.
Here, the example application is Wonderland. BattleTest is investigating an invoice API using accounts from two different organisations. The feed shows the agent's investigation and tool activity. The side panels keep the run structure and participants in view. The team can follow what is being tested while the assessment is running.
Maya adds context: billing moved onto a shared invoice API in the last release. Alex can reply in the same conversation. That gives the security engineer application context without moving the discussion to a separate chat. A message addressed to BattleTest asks the AI a question. Its answer appears in the same feed, alongside the team's messages.
An operator can also steer the investigation. Alex directs BattleTest to focus on the invoice API and check access between accounts. The directive appears in the feed, followed by BattleTest's response about the next test. A team message and an operator directive have different purposes: one adds context; the other directs the agent's work.
The finding shows that a request made with the Northstar account returned an invoice belonging to Juniper. Open the evidence to inspect the request and response. The account used for the request and the owner in the returned data make the access-control issue clear. Engineers can investigate the evidence with security colleagues and use it to work on a correction. The Room is included in BattleTest Enterprise.
Explore the product or book a conversation about your deployment.