Platform · Private beta
AiTM.
AI Threat Modeller
A downloadable, AI-assisted threat modelling tool built on our TRACE framework. It runs the full TRACE workflow on your specs and code and hands back a structured threat model to review. Coming soon, in private beta.
Desktop app for macOS, Windows, and Linux. Not yet publicly available.
What it is
TRACE, run by an agent, on your machine
AiTM takes the same methodology our researchers use and puts it in a desktop app. You feed it your sources, it works through the TRACE phases, and you stay in control at every step.
Your data stays with you
Projects and your encrypted API key live on your machine. Nothing is uploaded except the AI calls you choose to run.
OpenAI or Anthropic
Use your own model provider. The key is stored encrypted in your OS keychain and only ever sent to the provider you pick.
Built for any system
Domain profiles span generic IT, web, cloud, mobile, and ICS/OT as well as Web3, so the model fits whatever you're building.
The pipeline
Every TRACE phase, end to end
Run the phases one at a time and review each artifact, or kick off the full pipeline for a first pass. Optional phases toggle per project.
Scope & sources
Point it at your white papers, specs, architecture docs, and code. It builds the source inventory.
TRACE model
Actors, roles, assets, critical invariants, and edges, extracted and structured for review.
STRIDE
Threat identification and ranking across each component and flow.
Attack trees
The top-ranked threats decomposed into concrete, testable attack paths.
Collusion & coordination
Quorum, delegation, and multi-actor failure modes single-threat analysis misses.
Report
A prioritised threat model and roadmap you can hand to your team.
Request beta access
Tell us about your project and we will get back to you within one business day.