bug-hunt

Performs adversarial multi-agent code reviews to identify and triage defects.

3|1|Updated Feb 20, 2026
One-click install
npx skills add https://github.com/PavelLizunov/VPNRouter --skill bug-hunt-pavellizunov
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: bug-hunt
Source: https://github.com/PavelLizunov/VPNRouter/tree/main/.claude/skills/bug-hunt
Command: npx skills add https://github.com/PavelLizunov/VPNRouter --skill bug-hunt-pavellizunov

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill addresses the critical gap between code development and production stability by formalizing adversarial reviews to catch defects before they reach users.

Core Features & Use Cases

  • Adversarial Multi-Agent Review: Orchestrates multiple independent agents to audit code from distinct perspectives like security, concurrency, and correctness.
  • Defect Ledger Integration: Automatically gates stable releases by enforcing the resolution of P0/P1 defects tracked in the project's ledger.
  • Use Case: Before merging a complex refactor of the VPN routing logic, use this Skill to perform a multi-lens audit to ensure no regressions or security vulnerabilities are introduced.

Quick Start

Run the bug-hunt skill to perform an adversarial review of the current git diff against the project invariants.

Frequently Asked Questions about bug-hunt

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I perform an adversarial code review on a git diff to catch security vulnerabilities?

Adversarial code review audits a git diff by fanning out independent agents to verify code against defined invariants from distinct perspectives like security and correctness. Surviving defects are recorded in a centralized defect ledger for triage.

What is multi-agent bug hunting and how does it identify code regressions?

Multi-agent bug hunting orchestrates independent agents to audit code from distinct perspectives such as concurrency and correctness. It identifies regressions by verifying code against defined invariants and recording surviving defects in a centralized ledger.

Can I gate stable releases based on the resolution of tracked P0 and P1 defects?

You can gate stable releases by enforcing the resolution of P0 and P1 defects tracked in the defect ledger. This ensures production stability by preventing stable cuts until identified critical issues are fully resolved.

How do I audit a complex subsystem refactor for concurrency and stability issues?

Auditing a complex subsystem refactor involves fanning out independent agents to verify code against defined invariants from distinct perspectives like concurrency and stability. Identified defects are recorded in a centralized defect ledger for triage.

When do I need an adversarial multi-agent code audit for my project?

An adversarial multi-agent code audit is needed before merging complex refactors to ensure production stability. It formalizes the review process to catch security vulnerabilities and regressions by verifying code against project invariants.

Does this multi-agent code review approach work with existing project invariants?

This multi-agent code review approach works directly with existing project invariants by fanning out independent agents to verify code against them. Surviving defects are then recorded in a centralized defect ledger for resolution.