djtjwillia

Dogfiles — Development skill for Claude Code

Development community

Development environment as code: automated macOS setup + codified Claude Code skills, agents, and config.

How to install Dogfiles

This entry records only its repository, not the path inside it, so there is no exact command to give. Open djtjwillia/dogfiles and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Dogfiles does

Development environment as code: automated macOS setup + codified Claude Code skills, agents, and config.

Alternatives in Development

Claude Coder — Kodu is an autonomous coding agent that lives in your IDE 5.2k ★
Param Discover — Discover hidden HTTP parameters on a URL or list of URLs using Arjun (or x8 fallback) 4.5k ★
Claude Scholar — Semi-automated research assistant for academic research and software development 3k ★

README

dogfiles

this cowdog has dotfiles!

Development environment as code. A single `task init` provisions a macOS workstation — Homebrew packages, shell, git identities, terminal — **and** a fully codified [Claude Code](https://docs.anthropic.com/en/docs/claude-code) setup. Both are managed with the same rigor as production infrastructure: idempotent, rendered from a canonical source, reviewable in git, and dry-run-able before anything touches the machine.

The repo is the source of truth. Deployed destinations (`~/.claude`, `~/.dotfiles`, `~/.gitconfig`, …) are outputs — never hand-edited.

Claude Code setup

The `claude/` directory is a version-controlled Claude Code configuration, synced to `~/.claude` by task targets. It treats an agentic coding workflow as infrastructure worth codifying and reviewing.

**What lives in `claude/`:**

CLAUDE.md + charter-details.md — a structured multi-agent operating model. The core rules load on every session; the detail file is referenced on demand to keep context lean. The model is explicit about how the assistant behaves: plan by default (no file edits until the user promotes the session), proactive routing to specialist subagents by domain, staged write permissions (plan → probe → narrow → wide), and output gates requiring verification steps and a rollback path on every plan. (It is themed as a "Synod Council" persona; the substance underneath is the operating model, not the theme.)
agents/ — 12 specialist subagent definitions (synod-*.md): architecture, security, data safety, CI/CD, docs/planning, debugging, code review, DX, dependency currency, UX, implementation, and routing. Each carries structured frontmatter with routing triggers and write/veto scope. Plus an eval/ directory for scenario-based routing evals.
commands/ — custom slash commands (run-evals, summary).
skills/ — vendored, customized skills: handoff, obsidian-summary, and obsidian-transcript.