ZacheryGlass .claude workflow
A personal ~/.claude for Claude Code: twelve subagents, review commands, worktree and goal skills and a Go status line.
- 602 ★ on the repository
- Last commit
Extended access
Why this level
- settings.json allows a wide range of read commands, Git and requests to any domain without prompting.
- The repository carries prebuilt status line binaries that run on every refresh.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add workflows/zacheryglass-claudeDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- settings.json allows a wide range of read commands, Git and requests to any domain without prompting.
- The repository carries prebuilt status line binaries that run on every refresh.
This is third-party code. Review the repository files before installing.
What it does
The repository holds one developer's working Claude Code configuration that can be copied into `~/.claude/` whole or in parts. `settings.json` sets allowed commands and hooks, and `CLAUDE.md` holds minimal agent instructions. There are six skills. `gemini-agent` hands a task to Gemini CLI for a second opinion or token savings, `generate-goal` turns a conversation into a self-contained `/goal` prompt, `run-goals` runs several such prompts in parallel as non-interactive Claude Code sessions, each in its own git worktree, `git-worktree-prune` deletes stale worktrees and branches while keeping those holding work found nowhere else, `root-cause-tracing` traces a bug backward through the call stack with instrumentation, and `skill-creator` scaffolds and refines new skills. Commands launch reviews: `/arch-review`, `/bugs`, `/perf-check`, `/ui-review`, `/docs`, `/arewedone` for a completeness check and `/review-gauntlet`, which runs all standard reviews in parallel on current changes and then addresses the findings. The twelve subagents include an architecture reviewer, a logic-error hunter, a C safety reviewer and an interface compatibility reviewer. A program in Go draws the status line: project, model with effort level, cache tokens, a 60-minute cache timer, Git branch and remaining context.
Who it is for. Claude Code users assembling their own set of subagents, review commands and hooks who want a working example to borrow from.
Good fit when
- You want a ready set of subagents for architecture, performance, UI and documentation review.
- You want to launch several goals in separate worktrees at once and get a summary of results.
- Old worktrees and branches have piled up and must go without losing the only copy of any work.
- You need a compact status line with a cache timer and remaining context.
- You want a settings.json sample with permissions and hooks as a base for your own.
Not a fit when
- You need an install package or a plugin: the repository is a personal directory that is copied by hand.
- Your platform is neither Windows nor macOS: the status line is built for those two, while the rest works anywhere.
- The repository has no git: the worktree and goal skills rely on git worktree.
Example request
Run /review-gauntlet on my current changes and fix the issues it finds.Limitations
The repository is not packaged as a plugin. Files are copied into `~/.claude/` by hand and the settings adjusted for your environment.
The status line is built for Windows and macOS: the repository carries ready binaries `statusline.exe`, `statusline-darwin-amd64` and `statusline-darwin-arm64`, with build scripts for both systems. Binaries from someone else's repository are better rebuilt from `statusline.go`.
The `github_issue_guard.py` and `protect_claude_md.py` hooks are not enabled in `settings.json`. The `gemini-agent` skill needs Gemini CLI installed.
How to disable. Delete the copied files from `~/.claude/` and restore your previous `settings.json`, then restart Claude Code.
FAQ
How do I use this configuration?
Clone the repository, copy the files you want into `~/.claude/`, adjust permissions to taste and restart Claude Code.
What does /review-gauntlet do?
It runs all standard subagent reviews in parallel on current changes, then addresses the findings.
How does the run-goals skill run goals?
Each `/goal` goes into its own non-interactive Claude Code session in its own git worktree, after which the skill gathers a results report.
What does the status line show?
Project, model with effort level, cache read and write tokens, a 60-minute cache timer, the Git branch with status and the percent of context left, color-coded.
Is settings.json safe for someone else's environment?
No. It is written for personal work and allows a wide range of read commands, Git and domains without prompting. Read it line by line and keep only what you need.
Discussion
No questions or reviews yet.
Related
WorkflowsA personal Claude Code setup: four loading layers, an advisor model mode, ten hooks and working commands
Claudette runs several Claude Code sessions in parallel, each in its own git worktree, from a window or the console
Workflows for Claude Code: /do, /omo, /bmad-pilot and more commands, with Codex, Gemini or OpenCode writing the code through codeagent-wrapper
A Claude Code workspace folder: hooks with a local Ollama detect the project type, pick skills and remind you to update the docs
claude-mods adds 112 skills, 3 agents and 3 commands to Claude Code: stacks, agent operation, defense, session continuity.
A template for Claude Code projects: parallel agent swarms, workflow skills, safety hooks and rules in one install.
README in short
The README lists the repository contents by folder: settings, status line, CLAUDE.md, hooks, skills, commands and subagents. A separate section describes the status line with a performance table and segments, followed by official documentation links and four usage steps.