perseus:audit

Vulnerability analysis traces untrusted data from sources to sinks by applying a negative analysis loop that validates defenses at each step, verifying potential exploit paths across software components to surface exploitable weaknesses.

Analyzing code for XSS and injection vulnerabilities is best handled through parallel analysis across dedicated agent types, tracing untrusted data flows to validate defenses and uncover potential exploit paths within software components.

Yes, you can use a parallel-agent workflow to audit authentication, authorization, and SSRF vulnerabilities, with each agent type producing dedicated deliverables and contributing to an exploit queue for verification.

Generating an exploit queue for proof of concept development involves running a parallel vulnerability analysis across components, tracing untrusted data flows to sinks, and outputting per-agent deliverables that reveal exploitable paths.

Yes, you need to identify components beforehand, as this vulnerability analysis operates during Phase 3 by analyzing components identified in Phase 1 to uncover potential exploit paths and weaknesses.

White-box vulnerability analysis requires validating defenses at each step because a negative analysis loop traces data flow from untrusted sources to sinks, ensuring all potential exploit paths are exposed before generating actionable insights.