My Dotfiles

Overview

This repository contains my personal dotfiles, managed using chezmoi. It sets up my development environment, including configurations for Zsh, Neovim, Git, and various development tools. Tool versions are managed using mise-en-place.

Installation

1. Install chezmoi: Follow the instructions on the chezmoi installation guide.
2. Initialize chezmoi with this repository:

   chezmoi init https://github.com/laurigates/dotfiles.git

3. Review the changes: Check which files chezmoi plans to create or modify.

   chezmoi diff

4. Apply the changes:

   chezmoi apply -v

Tool Management with mise-en-place

This setup uses mise-en-place (formerly rtx) to manage development tool versions (like Node.js, Python, Go, etc.).

• Tool versions are defined in the .config/mise/config.toml file (managed by chezmoi).
• After cloning or updating the dotfiles, run mise install in your shell to install the specified tool versions.
• mise automatically activates the correct tool versions when you enter a directory containing a mise.toml or .tool-versions file.

Claude Code Plugins 🔌

This repository now includes Claude Code plugins for easy installation and management of AI agents and slash commands!

Quick Start

# Add this repository as a plugin marketplace
/plugin marketplace add laurigates/dotfiles

# Install the dotfiles toolkit plugin
/plugin install dotfiles-toolkit

What you get: 30+ specialized agents and 20+ slash commands for development, infrastructure, documentation, and more.

📖 Full guide: See PLUGINS.md for complete documentation.

AI Tools & MCP Configuration

AI tools and MCP (Model Context Protocol) servers are configured through the .chezmoidata.toml file and automatically installed via the update-ai-tools.sh script.

MCP Server Configuration

MCP servers for Claude Code are dynamically configured from .chezmoidata.toml. To manage servers:

• Enable/disable servers: Set enabled = true/false in the [mcp_servers] section
• Add new servers: Create a new [mcp_servers.name] section with required fields
• Configure options: scope, command, args, and optional transport

Example configuration:

[mcp_servers.my-server]
  enabled = true
  scope = "user"
  command = "npx"
  args = ["-y", "my-mcp-package"]
  transport = "stdio"  # optional

Adding/updating servers (safe during active Claude sessions):

./update-ai-tools.sh  # or: chezmoi apply update-ai-tools.sh

Cleaning up disabled servers (WARNING: disrupts active Claude sessions):

./cleanup-mcp-servers.sh  # Run only when no Claude sessions are active

Note: update-ai-tools.sh only adds new servers without removing existing ones, making it safe to run during active Claude sessions. Use cleanup-mcp-servers.sh only when you need to remove disabled servers and no Claude Code sessions are running.

Claude Code Configuration

Plugins

The repository provides three Claude Code plugins via a local marketplace for instant development feedback:

# Plugins auto-load from: ~/.local/share/chezmoi/plugins/
# Check available plugins:
/plugin list

Available plugins:

• dotfiles-core: Essential development workflows (git, code review, testing, docs, CI/CD)
• dotfiles-experimental: Testing new automation features (devloop, modernization) - can disable if unstable
• dotfiles-fvh: Work-specific Podio integrations - disable for personal projects

Enable/Disable in settings:

"enabledPlugins": {
  "dotfiles-core@dotfiles": true,
  "dotfiles-experimental@dotfiles": true,  // Set to false to disable
  "dotfiles-fvh@dotfiles": false           // Disabled for personal use
}

Development workflow:

# Edit plugin/skill files directly in source directory
vim ~/.local/share/chezmoi/.claude/skills/python-development/SKILL.md
vim ~/.local/share/chezmoi/plugins/dotfiles-core/commands/git/git:commit.md

# Apply changes to ~/.claude after editing
chezmoi apply -v ~/.claude  # Or use alias: ca-claude

Why exact_dot_claude/ instead of symlink? The .claude directory uses chezmoi's exact_ prefix for atomic updates and auto-cleanup of orphaned files. This prevents race conditions with running Claude processes and ensures predictable state. Run chezmoi apply -v ~/.claude after editing skills or commands.

Full guide: See PLUGINS.md and CLAUDE.md

Skills

32 auto-discovered skills provide contextual guidance:

• Core Tools: chezmoi-expert, shell-expert, fd-file-finding, rg-code-search, jq-json-processing, yq-yaml-processing, ast-grep-search, vectorcode-search
• Version Control: git-branch-pr-workflow, git-commit-workflow, git-security-checks, git-repo-detection, release-please-protection
• GitHub Actions: claude-code-github-workflows, github-actions-mcp-config, github-actions-auth-security, github-actions-inspection
• Languages: python-development, rust-development, nodejs-development, cpp-development, neovim-configuration
• Infrastructure: container-development, kubernetes-operations, infrastructure-terraform, embedded-systems
• Agent/Memory: agent-coordination-patterns, agent-file-coordination, multi-agent-workflows, graphiti-episode-storage, graphiti-learning-workflows, graphiti-memory-retrieval

Skills activate automatically based on your work context. See CLAUDE.md for the skills catalog.

Further Documentation

Components Overview

For a detailed breakdown of the tools and components managed by these dotfiles (like chezmoi, mise, Zsh plugins, Neovim setup, etc.), see:

• Components and Workflow Tools

Further Documentation

For more specific guides, see the following documents:

• Platform Specific Notes
• Container Testing
• Debugging Guide
• Verifying Environment Setup