claude-code-agent

Delegate coding and review tasks to Claude Code CLI sessions with structured JSON outputs.

Updated Mar 12, 2026
One-click install
npx skills add https://github.com/Ben2pc/g-claude-code-plugins --skill claude-code-agent-ben2pc
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: claude-code-agent
Source: https://github.com/Ben2pc/g-claude-code-plugins/tree/main/skills/claude-code-agent
Command: npx skills add https://github.com/Ben2pc/g-claude-code-plugins --skill claude-code-agent-ben2pc

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

The skill enables delegating coding, review, diagnosis, planning, and structured output tasks to independent Claude Code sessions via the Claude Code CLI entry point.

Core Features & Use Cases

  • Handoff to autonomous Claude Code sessions for execution and review.
  • Supports multi-turn conversations with --resume / --continue, and returns structured outputs suitable for automation.
  • Works in scripted or CI contexts with --worktree, tool whitelists, and explicit --output-format control.

Quick Start

Start a sample task in your project directory with claude -p 'Describe the task' --output-format json to begin a session and receive a structured result.

Frequently Asked Questions about claude-code-agent

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

FAQPage Schema
How do I delegate coding tasks to an autonomous Claude Code session?

You can delegate coding tasks by running the Claude Code CLI with a prompt like claude -p 'Describe the task' --output-format json, which starts an independent session and returns a structured result.

Can I resume multi-turn coding conversations in a scripted or CI workflow?

Yes, you can resume multi-turn coding conversations in CI workflows by using the --resume or --continue flags to maintain context across independent Claude Code sessions.

How do I isolate automated code review tasks using Claude Code?

You can isolate automated code review tasks by launching Claude Code with --worktree and --bare flags, ensuring the session operates in an isolated environment separate from your main directory.

Does the Claude Code CLI support structured JSON outputs for automation?

Yes, the Claude Code CLI supports structured JSON outputs through the --output-format flag, making it suitable for automation and scripted workflows that require explicit machine-readable formats.

What is the best way to restrict tool access during autonomous coding sessions?

The best way to restrict tool access during autonomous coding sessions is by applying tool whitelists when delegating tasks, ensuring the Claude Code agent only uses explicitly permitted tools.

Do I need an authenticated session to run Claude Code for diagnostics?

Yes, you need the Claude Code CLI installed and an authenticated session active before delegating diagnostic, planning, or coding tasks to independent agent sessions.