vigolium-audit

What problem does it solve?

It helps you find exploitable security issues in a repository and turn noisy analysis into a structured, confidence-improving findings tree without manually coordinating a complex audit workflow.

Core Features & Use Cases

• Autonomous multi-agent audit pipeline: Runs a fixed methodology (intel → SAST → adversarial review → PoC → report) to reduce false positives.
• Mode-based depth control: Supports lite/balanced/deep plus follow-ups like revisit, reinvest, confirm, diff, merge, and longshot to match your intent and constraints.
• Resume and machine-readable outputs: Continues interrupted runs and emits structured JSON (NDJSON streams) for automation.

Example use case: You need to assess a newly deployed service for critical and high-impact vulnerabilities and want a prioritized, finalized report with PoCs and a stable findings structure.

Quick Start

Run a full deep audit with Claude against your repository by executing: vigolium-audit run --mode deep --agent claude --target /path/to/repo.