mantis-summarize

Generate security-focused directory summaries via bottom-up repository traversal.

731|84|Updated Jun 15, 2026
One-click install
npx skills add https://github.com/google/mantis --skill mantis-summarize
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: mantis-summarize
Source: https://github.com/google/mantis/tree/main/mantis-summarize
Command: npx skills add https://github.com/google/mantis --skill mantis-summarize

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This skill addresses the challenge of high token overhead during security reviews by generating concise, security-focused summaries of directory contents, allowing agents to navigate large codebases efficiently.

Core Features & Use Cases

  • Automated Repository Mapping: Recursively generates security-focused mantis-summary.md files for directory structures.
  • Context Optimization: Reduces token usage by abstracting low-level details while preserving critical security information like trust boundaries and sensitive operations.
  • Use Case: When starting a new security audit on a large, unfamiliar repository, use this skill to create a high-level map of the codebase to guide your threat modeling and planning phases.

Quick Start

Run the mantis-summarize command to generate security-focused summaries for all directories in the current repository.

Frequently Asked Questions about mantis-summarize

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

FAQPage Schema
How do I map a large codebase for a security review without exceeding token limits?

Security-focused codebase mapping solves high token overhead by generating concise summaries of directory contents. It performs a bottom-up traversal to aggregate component, API, and vulnerability data, allowing agents to navigate large repositories efficiently during security analysis.

What is the best way to generate security summaries for repository directory trees?

To generate security summaries for repository directory trees, run an automated command that recursively creates security-focused summary files. It operates by performing a bottom-up traversal of the repository tree to aggregate component, API, and vulnerability data into deterministic, actionable codebase maps.

Do I need local filesystem access to generate codebase maps for threat modeling?

Yes, local filesystem access is required to generate codebase maps for threat modeling. The process requires local filesystem access and historical learning logs to produce deterministic, actionable repository maps and security-focused directory summaries.

Can I use automated repository mapping to optimize context window usage for security analysis?

Yes, automated repository mapping optimizes context window usage for security analysis by abstracting low-level details into concise directory summaries. It reduces token usage while preserving critical security information like trust boundaries and sensitive operations for downstream threat modeling.

What limitations should I expect when using automated codebase summaries for security audits?

Limitations of automated codebase summaries for security audits include the requirement for local filesystem access and historical learning logs. Additionally, abstracting low-level details to optimize token usage may omit granular implementation logic needed for highly specific vulnerability investigations.