agent-development

Design Claude Code agents with declarative prompts and standardized tool access.

18|4|Updated Jan 15, 2026

Ready to run

npx skills add https://github.com/ataschz/tanstack-start-mastra-example --skill agent-development-ataschz

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: agent-development
Source: https://github.com/ataschz/tanstack-start-mastra-example/tree/main/.agents/skills/agent-development
Command: npx skills add https://github.com/ataschz/tanstack-start-mastra-example --skill agent-development-ataschz

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps teams design and maintain robust Claude Code agents by providing a consistent pattern for agent prompts, memory management, and delegation strategies, reducing setup time and errors.

Core Features & Use Cases

Declarative prompt pattern: define role, triggers, and delegation behavior to ensure automatic task routing.
Memory and session guidance: establish memory limits, recovery steps, and session restart requirements.
Pipeline orchestration: manage workflow positions and Next links to maintain ordered agent pipelines.
Tool access standardization: provide a unified toolset across agents to simplify configuration.

Quick Start

To begin, create or update an agent definition under .claude/agents with a proper frontmatter and a description following the Declarative Over Imperative principle. Restart your Claude Code session to load the new agent. Optionally review and adjust the Next field in the agent pipeline to ensure correct sequencing.

Frequently Asked Questions about agent-development

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

FAQPage Schema

How do I design Claude Code agents with smart delegation?▼

Design Claude Code agents by applying a declarative prompt pattern that defines roles, triggers, and delegation behavior, ensuring automatic task routing across agent pipelines.

How do I structure custom agent prompts for automatic task routing?▼

Structure custom agent prompts declaratively by specifying the agent role, triggers, and delegation behavior in the agent definition file under .claude/agents to ensure automatic task routing.

How do I manage memory limits and session restoration for Claude Code agents?▼

Manage memory limits and session restoration by establishing recovery steps, memory limits, and session restart requirements within the agent definition to maintain robust agent pipelines.

What's the best way to orchestrate ordered agent pipelines in Claude Code?▼

Orchestrate ordered agent pipelines by managing workflow positions and configuring the Next field in each agent definition to maintain correct sequencing across the pipeline.

Do I need to restart the Claude Code session to load a new agent?▼

Yes, you need to restart your Claude Code session after creating or updating an agent definition under .claude/agents to load the new agent configuration properly.

Can I standardize tool access across multiple Claude Code agents?▼

Yes, you can provide a unified toolset across agents to simplify configuration, applying standardized tool access patterns and self-documenting prompts within your Claude Code projects.