superagent_

// products

Attack scenarios

[ view markdown ]

Each attack scenario focuses a Red Teaming report on one security question. Browse 49 built-in scenarios to see what they check and what a finding can tell you. Each page links to related OWASP, CWE, or vendor guidance.

Applications · Agents · Repositories

Choose a scenario

Choose Attack scenario in red-team settings or when creating a report. General assessment is the default and keeps the full methodology. Search the dropdown by name, objective, or ID to select a focused scenario for your target category.

A selected scenario focuses the assessment on that objective. Unselected scenarios are outside the assessment and must not be treated as tested. For Agents, unrelated framework entries are marked outside_selected_assessment.

Supply the prerequisites

Review the scenario’s required capabilities and setup. Supply authorized identities, expected policy, and disposable fixtures needed to exercise the objective. Keep passwords and tokens in the encrypted credential fields. Use the specific prompt for non-secret fixture references and expected policy.

Missing prerequisites are reported as blocked. Unsupported capabilities require a not applicable explanation. Selecting a scenario does not provide missing capabilities or expand the authorized scope.

Create and manage custom scenarios

Organization owners can choose Add new scenario in the dropdown or Edit scenario for a selected custom scenario. The modal requires only a Name and What should we test? in plain language. Selecting a preset first prefills the instructions with its objective. Superagent discovers the relevant capabilities and determines the testing approach, evidence, and finding classification; missing required setup remains blocked.

The Custom scenarios section in the Red teaming tab lists reusable scenarios for the target category. Owners can add, edit, or delete them there. Each save creates a revision. Settings apply the selected revision to manual and scheduled runs; existing reports and reruns preserve their resolved definition. Deleting removes a scenario from the library, while existing reports and saved run settings retain their pinned version.

Interpret scenario coverage

The report labels the selected scenario and tracks coverage separately from findings:

Coverage Meaning
Verified vulnerability A linked finding passed the existing verification gates.
Blocked Required setup or prerequisites were unavailable.
Not applicable The target does not support a capability required by the scenario; the report explains why.
Inconclusive The available evidence does not establish a verified outcome.
Not tested The scenario has not been exercised.

Runner-reported negative results remain inconclusive until independently verified outside the target environment, even when baseline, denied-control, and attack-variant evidence is supplied. Zero findings alone is not a pass. Scope, verification, peer review, and remediation requirements still apply. Proposed control relationships are scoped evidence references and do not establish certification or full compliance.

Applications

Agents

Repositories