Developer Builds AI-Assisted Security CLI Tool for OWASP BLT Using Rust and MCP

·1 views

A developer has completed a summer project building two integrated tools for OWASP BLT: BLT-MCP, which connects AI models to the BLT API, and BLT-CLI, a Rust-based terminal interface for security work. The CLI allows users to browse vulnerabilities, search reports, view bounty hunts, submit findings, and access AI-powered analysis — all from the terminal. An AI triage feature can assess severity, attack vectors, and duplicate risk, but deliberately stops short of making automated changes, leaving final decisions to the user. A vision model integration now lets users upload screenshots to help auto-fill vulnerability report fields such as title and reproduction steps. The developer notes the current release is a foundation, with planned improvements to AI analysis quality, error handling, and overall workflow reliability.

Read the full story at DEV Community

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

How Flutter Developers Can Build Reliable AI Features Using Structured Outputs and Streaming

A technical guide aimed at Flutter developers outlines three core skills required for production-ready AI engineering: structured outputs, real-time streaming, and robust error handling. The article explains why free-text AI responses can be unpredictable and difficult to manage in real applications, advocating instead for JSON-formatted structured outputs. Using Google's Gemini model via the googleai_dart client, the guide demonstrates how to enforce strict JSON schemas and deserialize responses into strongly typed Dart models. It also covers retry logic with backoff strategies and safety-filter exception handling to prevent silent failures. The series, now in its second part, targets Flutter developers looking to build cleaner, more enterprise-grade AI integrations.

0 comments Read more at DEV Community

ProgrammingHacker News ·

Civic Hygiene: The Case Against Building Tools That Enable Police States

A blog post titled 'Civic Hygiene,' originally published in November 2013 on shkspr.mobi, has resurfaced on Hacker News, attracting significant attention. The piece argues that technologists and developers have a moral responsibility to avoid creating tools that could be exploited by authoritarian or surveillance-heavy governments. The post encourages engineers to consider the potential misuse of their work before building it. It gained 96 upvotes and sparked 39 comments on Hacker News, indicating strong community interest in the ethical dimensions of technology development.

0 comments Read more at Hacker News

ProgrammingDEV Community ·

Techucation Launches Structured Quantum Computing Learning Path for 2026

Techucation has published a comprehensive quantum computing learning series titled 'Quantum Series 2026,' designed to guide learners from foundational concepts to advanced algorithms. The series covers topics ranging from qubits and Hadamard gates to Grover's algorithm and quantum teleportation, structured as a progressive learning path rather than in the order the posts were written. Each article is self-contained but collectively forms a continuous curriculum built on first principles. The series is developed using Qiskit 1.x and includes hands-on posts with practical implementations. Readers already familiar with basic quantum concepts can skip directly to the algorithms section, making the series accessible to both beginners and those with prior knowledge.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Developer Builds Sentinel-2 Vegetation Map of Kefalonia Using Free ESA Data

A developer working on a project called ARGOS successfully processed the first Sentinel-2 satellite pass over the Greek island of Kefalonia within 48 hours. Using freely available European Space Agency imagery updated every five days, the goal was to generate a baseline NDVI vegetation map of the entire island. The pipeline, built with Python, rasterio, xarray, and GDAL, involved roughly twelve steps and required converting data from UTM 34N to WGS84 for web mapping. The resulting 180-line script — 40 lines of which are error handling — filters clouds, calculates NDVI, and outputs a GeoTIFF for eventual ingestion into PostGIS. The next phase will layer wildfire risk data using slope, aspect, NDVI, and fuel type, though the developer acknowledges the model currently misclassifies Greek fir burn behaviour.

0 comments Read more at DEV Community