Bastion

Inspiration

The inspiration behind our project Bastion was the popular tv show Suits which had included high stakes cases where even a single mistake be it technology or the lawyer side could cause catastrophic losses from be it monetary wise or reputation.

What it does

Bastion is a cyber incident response workspace where teams can track affected systems, response tasks, and incident updates. Its Containment Lab lets responders simulate isolating a system and compare which attack paths would be cut, which systems could still be reached, and which services could go offline. Teams can save a proposed containment plan for review. The simulation runs on a model of the incident, so the team can explore options safely. We used Lovable to build the React and TypeScript interface, TanStack Start for the application, and Supabase for authentication and incident data. We built the Containment Lab around a graph with two kinds of connections. Attack paths represent where an attacker could move next, while service dependencies represent what each system needs to keep running. When a user selects a system to isolate, Bastion recalculates both outcomes and shows the results side by side.

Challenges we ran into

The hardest part we ran into during this creation of our project was the validation of the authenticating access code which would give access to a user of that specific incident.

Accomplishments that we're proud of

We built a demo that makes the trade-off visible in two clicks. In our fictional ransomware scenario, isolating the phished finance laptop keeps payroll running, but an attack path through the VPN still reaches it. Isolating the domain controller cuts the attack path to payroll, but payroll goes offline because it depends on that controller. Backups are still reachable through the file server. Seeing both outcomes gives the team something concrete to discuss before choosing a response.

What we learned

We learned that whether a system is reachable by an attacker and whether it is available to employees are separate questions. Showing the reason behind each outcome matters as much as showing the outcome itself. A responder can use that explanation to check the model's assumptions and explain the proposed action to the rest of the team.

What's next for Bastion

Something next for Bastion that we will be working on are ways that they can expand more than the law firm. For example first responders since there data is very valuable it is necessary for it to be safe system. Along with the genreal ways we would like to expand another way is increase the usability of anyone such as adding voice automated results which would decrease the need of tehcnology knowledge.