claude

Delegate code prompts to Anthropic's Claude Code CLI and evaluate responses.

1|Updated Mar 24, 2026
One-click install
npx skills add https://github.com/joesys/joesys-skills --skill claude-joesys
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: claude
Source: https://github.com/joesys/joesys-skills/tree/main/skills/claude
Command: npx skills add https://github.com/joesys/joesys-skills --skill claude-joesys

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Delegates code-centric prompts to Claude Code CLI and ensures a rigorous critical evaluation of the delegated output, so you always see the model's reasoning with accountability.

Core Features & Use Cases

  • Delegate prompts to Claude Code CLI for code analysis, refactoring, and automated editing.
  • Support session resume with named or anonymous sessions, and perform a critical evaluation after the delegated response.
  • Follow a strict post-delegation flow: present Claude's response verbatim, then provide critical evaluation and a concise summary.

Quick Start

Delegate a code prompt to Claude Code CLI and review Claude's evaluation.

Frequently Asked Questions about claude

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

FAQPage Schema
How do I delegate code analysis and refactoring tasks to Claude Code CLI?

You can delegate code analysis and refactoring to Claude Code CLI by sending your code-centric prompts through the delegation protocol. The Skill forwards your request, presents Claude's response verbatim, and follows up with a critical evaluation.

Can I resume a previous session when delegating prompts to Claude Code CLI?

Yes, you can resume a previous session when delegating prompts to Claude Code CLI. The Skill supports session resume functionality with both named and unnamed sessions, allowing you to continue prior code analysis or refactoring workflows.

How does critical evaluation work after Claude generates code analysis?

Critical evaluation works by reviewing Claude's generated code analysis for blind spots after the delegation completes. The Skill presents the verbatim response first, then provides a critical assessment and a concise summary to ensure accountability.

Does the Claude delegation Skill modify my project files automatically?

No, the Claude delegation Skill does not modify your project files automatically. It follows a strict running protocol that presents responses and evaluations without saving delegated outputs or altering your project unless explicitly requested.

What is the best way to evaluate blind spots in automated code editing responses?

The best way to evaluate blind spots in automated code editing responses is using a post-delegation critical evaluation flow. This approach presents the model's verbatim output first, then applies a critical assessment to identify reasoning gaps.

When should I not use prompt delegation for code refactoring tasks?

You should not use prompt delegation for code refactoring tasks when you need direct inline editing or unattended automated file modifications. The Skill is designed for delegate-and-evaluate workflows and intentionally avoids modifying your project unless requested.