TriLane
An autonomous AI pentesting tool that runs a staged, evidence-backed security audit of web apps and APIs across identity, injection, business-logic and configuration risks, for targets the user is authorized to test.
TriLane turns a single natural-language objective into a staged security audit. It first builds an attack-surface graph of routes, code paths and auth boundaries, then splits review across six semantic lanes covering identity and auth, injection, ingress and SSRF, business logic, configuration and secrets, and edge-case coverage, before merging duplicate claims into adjudicated findings with severity, evidence and payloads.
It runs as a desktop application built with Rust and Tauri, tracking scan, agent, findings and configuration state in a GUI, and offers a Safe mode for lower-risk exploration and a Lab mode that allows more active probing against explicitly authorized targets. Run transcripts are archived locally for regression analysis.