Building an AI-Powered Security TUI with Rust and MCP
Picking up where I left off In my midterm update , I introduced the two main parts I was working on for OWASP BLT: BLT-MCP , which connects AI models with the BLT API, and BLT-CLI , a Rust-based terminal interface for working with BLT. At that point, most of the foundation was there, but some of the features were still being worked on. The triage workflow was still coming together, and the…
The OWASP BLT project has made significant progress in developing an artificial intelligence-powered security terminal user interface (TUI) with Rust and MCP components. The BLT-CLI, a Rust-based terminal interface, now includes a range of features such as browsing vulnerabilities, searching and filtering reports, viewing bounty hunts, checking contributor rankings, inspecting individual vulnerabilities, submitting new reports, and utilizing AI for analysis and screenshots.
Guest mode is also available, allowing users to explore the application without prior configuration.
The BLT-CLI operates in two modes: basic mode, which communicates directly with the BLT API, and AI-assisted mode, which leverages the BLT-MCP server to connect the CLI with AI models. This separation allows the terminal interface to focus on user experience while the MCP handles the connection between services.
One of the key features being experimented with is AI-assisted triage. When examining a vulnerability, the AI can analyze available information and provide suggestions for severity, attack vectors, potential impact, duplicate risk, and other relevant considerations. The AI provides an analysis for review, ensuring that changes are not made automatically behind the scenes.
Additionally, the report submission workflow has been integrated into the CLI, enabling users to fill out report details such as title, URL, severity, bug type, description, and proof of concept directly from the terminal and submit it to BLT.
The project's vision component has also advanced, allowing users to provide a screenshot and have the vision model analyze it to help fill in parts of a vulnerability report, including the title and reproduction steps. The project is still evolving, with plans to improve AI analysis, MCP tools, TUI workflows, error handling, and add more functionality to the CLI. The author acknowledges ongoing experimentation in integrating AI into security workflows and emphasizes the project's status as a starting point.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
