cavecrew

Delegate code investigation, targeted edits, and diff reviews to specialized subagents.

Updated Feb 24, 2026

Search Tags:#code review #context management #subagent #code editing #code investigation

Ready to run

npx skills add https://github.com/VXerys/marketiv-web --skill cavecrew-vxerys

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: cavecrew
Source: https://github.com/VXerys/marketiv-web/tree/main/.qwen/skills/cavecrew
Command: npx skills add https://github.com/VXerys/marketiv-web --skill cavecrew-vxerys

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires haiku, sonnet, opus, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill helps streamline code investigation and editing by delegating specific tasks to subagents, reducing context exhaustion and improving efficiency.

Core Features & Use Cases

Code Investigation: Locate code using cavecrew-investigator for quick symbol lookup and dependency analysis.
Surgical Editing: Perform targeted edits on up to two files with cavecrew-builder.
Code Review: Conduct efficient diff reviews with cavecrew-reviewer.
Use Case: When you need to quickly find a specific function or module, use cavecrew-investigator. For a targeted edit, delegate to cavecrew-builder, and for a review, use cavecrew-reviewer.

Quick Start

Delegate code investigation to the cavecrew-investigator subagent for efficient symbol lookup.

Frequently Asked Questions about cavecrew

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

FAQPage Schema

How do I delegate code investigation tasks to subagents without exhausting context?▼

You can delegate code investigation to specialized subagents to prevent context exhaustion. This approach manages codebase navigation and symbol lookup efficiently, injecting results seamlessly into the main context.

What is the best way to perform surgical edits on specific files during a code review?▼

The best way to perform surgical edits is by delegating targeted file modifications to a builder subagent. It handles precise edits on up to two files and facilitates efficient diff reviews.

Can I use specific models like Haiku and Opus for different subagent tasks?▼

Yes, the system requires appropriate models for each subagent. Dependencies include Haiku, Sonnet, and Opus, allowing you to match model capabilities to specific code investigation and editing tasks.

How do I conduct an efficient diff review after targeted code editing?▼

You conduct an efficient diff review by delegating the task to a specialized reviewer subagent. This subagent focuses on reviewing the code changes, ensuring targeted edits meet quality standards.

When do I need subagent delegation for codebase navigation and dependency analysis?▼

You need subagent delegation for codebase navigation when you require quick symbol lookup and dependency analysis. Delegating these tasks prevents context exhaustion and improves overall investigation efficiency.

Does subagent-based code editing support multi-file modifications?▼

Subagent-based code editing supports targeted modifications on up to two files. It is designed for surgical edits rather than broad refactoring, ensuring precise changes within the codebase.