New
We just launched
N-Day-Bench
— a monthly benchmark for LLM vulnerability discovery
Learn more
Products
Solutions
Company
Resources
Research
Login
Book a Demo
CHANGELOG
What's new
The latest features, improvements, and updates to winfunc.
April 2026
N-Day-Bench
Public monthly benchmark measuring LLM vulnerability discovery across 1,000+ advisories
Open methodology with Curator, Finder, and Judge agent evaluation framework
Earlier
Software Composition Analysis
Detect vulnerable dependencies alongside SAST findings
Improved rendering for dependency vulnerability reports
Real-time event streaming for scan progress
AI Triager & Repository Rules
New agent that triages vulnerabilities and answers security questions in context
Customize agent behavior per-repository with configurable rule sets
Quick scan mode for faster, targeted scans
Secrets Detection
High-precision detection of API keys, tokens, passwords, and certificates in code and config files
Covers environment variables, hardcoded credentials, and private keys
API Security
Deep analysis of REST, GraphQL, and gRPC endpoints
Detects IDOR, broken authentication, missing rate limits, and injection vectors
Enterprise Controls
SSO/SAML support and role-based access management
Self-hosted deployment option with zero data retention
Repository-level access controls for admins
CI Integration & PR Security
Native pipeline integration for GitHub Actions, GitLab CI, Jenkins, and more
Scan every pull request with incremental diff-based analysis
Blocking gates, SARIF output, and inline PR comments
Custom Scan Rules
Configure focus and reporting rules to guide the AI agent’s analysis
Tailor scans to your security requirements and compliance needs
Real-Time Scanning
See findings as they’re discovered, not just after scan completion
Cancel running scans that are no longer needed
Scan job approval workflow
Function-Level Analysis
Deep code comprehension with reachability analysis and taint tracking
Complexity metrics and cross-reference mapping for every function
Analytics & Threat Hunter
Security score calculations, trend tracking, and token usage metrics
Specialized AI agent for deep vulnerability analysis
Query threats by specific file paths
Scanning Infrastructure
Background scan workers for reliable, scalable job execution
Directory ignore rules for repository analysis
Database connection pooling for improved performance
GitHub Integration
Connect repositories via GitHub App for seamless scanning
Streamlined OAuth authentication
Manual and automatic installation linking
Initial Release
AI-powered SAST with confidence scoring and complexity metrics
Enhanced context around vulnerable code paths
Real-time scan progress tracking and notifications
Vulnerability filtering, search, and navigation