Projects / YANOLINT
YanoLint
A web programming platform with an LLM-integrated language server that flags code vulnerabilities in real time.
- Date
- May 2026
- Type
- Capstone, BS Information Technology, University of Southeastern Philippines
- Team
- Mark David Prado, Raffy B. Suarez, Rovenado Nesta A. Villotes
- Adviser
- Ariel Roy L. Reyes
01 / Description
YanoLint is a web programming platform with an LLM-integrated language server that flags code vulnerabilities while you write code. It shows inline alerts, highlights the affected lines, and gives each finding a CWE label, a severity, an explanation, and a suggested fix. A Security Tutor answers follow-up questions about a finding.
It supports PHP, JavaScript, Java, and C, and focuses on five weakness types: path traversal (CWE-22), cross-site scripting (CWE-79), SQL injection (CWE-89), unrestricted file upload (CWE-434), and hard-coded credentials (CWE-798).
02 / How it works
- 01Web platform and dashboard for projects, files, and findings.
- 02Language Client and Language Server talk over the Language Server Protocol (LSP).
- 03An Analysis Component runs retrieval-augmented LLM analysis on a local model through Ollama and returns structured diagnostics.
- 04Built with TypeScript and Python, deployed with Docker.
- 05Continues the research in the PETI paper.
03 / Features
Real-time inline alerts in the editor
CWE label, severity, and explanation per finding
Suggested fix on hover
Vulnerabilities panel per project
Security Tutor for follow-up questions
Dashboard with an overview of findings