mantis

mantis

Provides security review capabilities for AI coding agents, automatically discovering, reproducing, and fixing vulnerabilities.

Agent SkillDevOpen source
Type
Agent Skill
Open source
Yes
GitHub Stars
★ 1.4k
Source
skill-github

Overview

Mantis is a modular, stack-agnostic security review toolkit designed to help AI coding agents autonomously discover, reproduce, and fix vulnerabilities. It works across various software or hardware stacks and can be adapted to specific domain needs. Using Mantis can enhance code security, but it should be handled with caution in isolated environments. Suitable for developers and teams with security requirements.

Capabilities

  • Automatically discover code vulnerabilities
  • Reproduce vulnerabilities to verify their existence
  • Generate and apply patches to fix vulnerabilities
  • Support multiple coding agent frameworks
  • Adapt to diverse domain-specific needs

Use cases

Code security reviewsAutomated vulnerability managementSecurity checks in continuous integrationDomain-specific security audits

Setup

Requires: API KeyNode 环境Docker
Install via CLI: npx skills add google/mantis

This information was compiled by AI from public sources and may contain inaccuracies — please refer to the source.

FAQ

Can Mantis be used in production environments?

Not recommended; should be run in isolated environments.

Which coding agent frameworks does Mantis support?

Supports multiple frameworks, such as Google ADK and Antigravity SDK.

Related skills