Why should I prefer Task sub-agents over remote API calls?

Task sub-agents have full access to local tools like Read, Grep, and Write, allowing them to reason across your entire codebase iteratively rather than being limited to a single API context window.

How can I fix memory issues when running multiple Claude Code agents?

You should increase the Node.js heap size by adding 'export NODE_OPTIONS="--max-old-space-size=16384"' to your shell profile and limiting parallel agent execution to 2-3 instances.

What is the benefit of limiting Bash tool access for sub-agents?

Limiting Bash prevents the model from using shell heredocs for file edits, which forces the use of optimized Write/Edit tools and significantly reduces the number of manual approval prompts.

How do I ensure Claude automatically delegates tasks to my custom agent?

Use a 'Strong Trigger Pattern' in the agent's description field, including explicit phrases like 'MUST BE USED when' and 'Use PROACTIVELY for' followed by specific task keywords.

Claude Code Agent Development Kit

Name: Claude Code Agent Development Kit
Author: brendadeeznuts1111

bybrendadeeznuts1111

0•

개발자 도구

Streamlines the design and optimization of custom Claude Code agents through effective delegation patterns and memory management.

This skill provides a comprehensive framework for building high-performance custom agents within the Claude Code ecosystem. It covers essential architectural patterns including auto-delegation triggers, granular tool access control, and declarative prompt engineering to ensure agents are both reliable and self-documenting. Developers can leverage this skill to solve common bottlenecks such as memory limits, repetitive manual approvals, and complex multi-agent pipeline orchestration, ultimately creating a more efficient and autonomous AI development environment.

주요 기능

01Granular tool permission patterns to minimize manual Bash approvals

02Pipeline management for numbered multi-agent workflows

03Declarative prompt framework for consistent agent outcomes

04Strong trigger patterns for reliable automatic agent delegation

05Node.js memory optimization and crash recovery configurations

060 GitHub stars

사용 사례

01Creating multi-stage agent pipelines with automated hand-offs

02Building specialized sub-agents for repetitive tasks like auditing or QA

03Optimizing agent prompts to handle large-scale code refactoring

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add brendadeeznuts1111/tier-1380-omega agent-development

For use in Claude.ai and ChatGPT

Download Skill