dev-setup-llm

Configure LLM agent setup for project repositories across Pi and Claude Code.

Updated Jun 17, 2025
One-click install
npx skills add https://github.com/nikbrunner/dots --skill dev-setup-llm
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: dev-setup-llm
Source: https://github.com/nikbrunner/dots/tree/main/common/.agents/skills/dev-setup-llm
Command: npx skills add https://github.com/nikbrunner/dots --skill dev-setup-llm

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Set up or maintain LLM agent configuration for a project repository, ensuring a scalable structure that supports fresh, legacy, and modern agent states.

Core Features & Use Cases

  • Tool-agnostic guidance for LLM agent setup across Pi and Claude Code environments.
  • Consolidates project-level AGENTS.md with global skill scaffolding, plus guides for Pi (guides/pi.md) and Claude Code (guides/claude-code.md).
  • Handles three states: fresh (no agent config), legacy (existing commands/agents), and modern (skills-based configurations) for streamlined onboarding and maintenance.

Quick Start

Create or update the AGENTS.md and .agents/skills/dev-setup-llm to reflect the project's LLM agent configuration using the provided guides.

Frequently Asked Questions about dev-setup-llm

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

FAQPage Schema
How do I configure an LLM agent setup for a new project repository?

To configure an LLM agent setup for a new project, initialize a fresh agent state by creating a centralized AGENTS.md file and scaffolding the .agents/skills/dev-setup-llm directory to establish tool-agnostic guidance for your repository.

Can I use this to maintain LLM agent configuration for legacy repositories?

Yes, you can maintain LLM agent configuration for legacy repositories by migrating existing commands and agents into a modern, skills-based configuration structure using the provided tool-specific guides.

What is the best way to set up a project configuration that supports both Pi and Claude Code?

The best way to support both Pi and Claude Code is applying a tool-agnostic workflow that consolidates project-level AGENTS.md configurations with specific setup guides located in guides/pi.md and guides/claude-code.md.

How does the AGENTS.md file work with the centralized skills directory?

The AGENTS.md file works by providing shared project-level agent instructions, while the centralized .agents/skills/dev-setup-llm directory delivers scalable scaffolding to apply configurations across different LLM agent environments.

Do I need separate configurations for Claude Code and Pi environments?

You do not need entirely separate configurations; the tool-agnostic workflow uses a centralized .agents/skills directory alongside dedicated guides for Pi and Claude Code to streamline onboarding across both environments.