Agentic Dotfiles

My development setup for working with coding agents across the terminal, editor, and shell.

It brings together shared instructions, reusable skills, specialized agent roles, client configuration for Codex, Claude Code, and Grok, plus the shell and tool settings that support the workflow. Use the repository as a starting point: adopt the pieces that fit how you work and change the rest to suit your machine and preferences.

Highlights

• Shared working principles across Codex, Claude Code, and Grok
• Specialized explorer, implementer, reviewer, and council agent roles
• Reusable skills for planning, implementation, review, documentation, and handoff
• Hooks that keep agent behavior consistent across a coding session
• Fish as the primary shell, with shared daily behavior mirrored into Zsh
• Terminal and editor configuration for a keyboard-driven workflow
• Small local utilities for project checks and agent-assisted development

Where to Start

If you want to adapt…	Start with
The overall agent workflow	AGENTS.md and .codex/AGENTS.md
Codex	.codex/config.toml, agents, and hooks
Claude Code	.claude/settings.json and agents
Grok	.grok/config.toml and hooks
Reusable agent skills	.skills-mgr/skills/
Fish	.config/fish/config.fish
Zsh	.zshrc and .zsh/fish-mirror.zsh
Bash	.bashrc
Editors and terminal tools	.config/

The CONTEXT-*.md files are maps for the less obvious parts of the setup. They explain which files own agent instructions, hooks, lifecycle behavior, and shared shell behavior.

Adapting the Agent Setup

Start with the instruction files before copying client settings. They define how agents are expected to explore a repository, make changes, validate their work, and communicate results.

The agent definitions divide work by responsibility:

• Explorers investigate ownership, behavior, and change impact.
• Implementers make focused changes after the scope is settled.
• Reviewers inspect correctness without owning the implementation.
• Council members provide independent judgment for genuinely ambiguous decisions.

Skills under .skills-mgr/skills/ provide task-specific workflows that can be shared by multiple agent clients. Begin with only the roles and skills you need; the setup is intentionally modular.

Before using the client configurations, review their models, permissions, enabled features, hooks, plugins, and external integrations. Some settings assume broad filesystem and command access because they are designed for a trusted local development environment.

Agent Skills

Only shared skills under .skills-mgr/skills/ and Codex skills directly under .codex/skills/ are listed below. “Model visible” means the model can select the skill itself when its condition is met; No skills are available only through an explicit workflow.

Name	Purpose	model visible	condition
assess-change-impact	Map callers affected by a shared change	Yes	Always
build-code-skeleton	Create an initial compile-safe project skeleton	Yes	Always
codebase-review	Review the whole working tree	No	Always
council	Gather independent agent judgments	Yes	Always
deliver-vertical-slice	Deliver an approved change end to end	Yes	Always
deslop	Reduce production code while preserving behavior	No	Always
dump-last-response	Save the preceding assistant response	No	Always
final-review	Review a completed delivery independently	Yes	Always
go-json-v2	Apply Go's encoding/json/v2 APIs	Yes	Go project
go-microoptimizations	Optimize measured Go hot paths	No	Go project
golang-best-practices	Apply modern Go practices	Yes	Go project
handoff	Prepare a compact handoff for another agent	No	Always
human-flavoured-writing	Write natural, human-sounding project copy	No	Always
js-ts-best-practices	Apply JavaScript and TypeScript practices	Yes	JavaScript or TypeScript project
juststore-rendering-optimizer	Reduce React rerenders with juststore	Yes	JavaScript or TypeScript project with juststore
new-project	Run the new-project workflow	Yes	Always
orchestrated-workflow	Coordinate a change through Codex agents	No	Always
postgres-17-18-features	Apply PostgreSQL 17 and 18 features	Yes	PostgreSQL project
read-codex-session	Inspect local Codex session transcripts	No	Always
session-usage	Report current Codex token usage	No	Always
shadowtree	Run and author Shadowtree recipes	Yes	Always
user-experience	Improve user-facing workflow behavior	Yes	Always
using-pjdoc	Validate indexed project documentation	Yes	Always
writing-readme	Write or improve repository READMEs	Yes	Always

Adapting the Shell Setup

Fish is the main shell configuration. Zsh loads a native port of the daily Fish behavior, while Bash has a smaller independent setup.

Do not replace your existing dotfiles wholesale. Compare each file with your current configuration and merge the parts you want. In particular, check:

• commands and plugins that may not be installed on your machine;
• Homebrew and other platform-specific paths;
• terminal capabilities, key bindings, and clipboard commands;
• editor, pager, history, prompt, and completion preferences;
• environment variables and local directory assumptions.

Keeping your existing configuration beside this repository makes it easier to adopt one layer at a time and roll back anything that does not fit.

Repository Map

.
├── AGENTS.md              # Working principles and repository guidance
├── .codex/                # Codex settings, agents, hooks, and skills
├── .claude/               # Claude Code settings and agents
├── .grok/                 # Grok settings and Codex-hook adapters
├── .skills-mgr/skills/    # Reusable cross-client skills
├── .config/fish/          # Primary shell configuration
├── .zshrc                 # Zsh-specific configuration
├── .zsh/                  # Shared behavior ported from Fish to Zsh
├── .bashrc                # Bash-specific configuration
├── .config/               # Selected editor, terminal, and CLI settings
└── .local/bin/            # Small development utilities

License

Licensed under the MIT License. You are welcome to copy, modify, and adapt the setup for your own workflow.