~ / rigs / witekbobrowski / dotfiles

witekbobrowski/dotfiles

🗂 Automated custom macOS config

↗ GitHub ★ 10 no license updated 1d ago personal setup Claude CodeCodex
share on X
ARCHETYPE
Orchestrator
A bench of specialised subagents. The main agent mostly delegates.
CONTEXT TAX · EVERY TURN
~988 tokens
Featherweight · median rig: 2.3k · breakdown
GUARDRAILS
0/5
No committed guardrails · details

Copy this rig

# review before running: this installs third-party code
$ npx degit witekbobrowski/dotfiles/agents ./rig-dotfiles/agents

MCP servers are added to Claude Code at local scope; env vars are shown as YOUR_… placeholders — we never store values. Files are fetched with degit into a separate folder so you can review before merging.

$ curl -fsSL --create-dirs -o .claude/agents/codex-reviewer.md https://raw.githubusercontent.<redacted>-reviewer.md
$ curl -fsSL --create-dirs -o .claude/agents/codex.md https://raw.githubusercontent.<redacted>.md
$ curl -fsSL --create-dirs -o .claude/agents/implementer.md https://raw.githubusercontent.<redacted>.md
$ curl -fsSL --create-dirs -o .claude/agents/researcher.md https://raw.githubusercontent.<redacted>.md
$ curl -fsSL --create-dirs -o .claude/agents/runner.md https://raw.githubusercontent.<redacted>.md
$ curl -fsSL --create-dirs -o .claude/agents/scout.md https://raw.githubusercontent.<redacted>.md

This rig commits no guardrails. Here is the community baseline instead — the deny/ask rules most often found across all 7,204 rigs:

{
  "permissions": {
    "deny": [
      "Read(./.env)",
      "Read(**/.env)",
      "Read(~/.ssh/**)",
      "Bash(rm -rf *)",
      "Read(**/*.pem)",
      "Bash(rm -rf /)",
      "Bash(git push --force:*)",
      "Bash(sudo *)",
      "Read(.env)",
      "Bash(rm -rf /*)",
      "Read(./.env.*)",
      "Read(~/.aws/**)",
      "Bash(git push --force*)",
      "Bash(rm -rf:*)",
      "Read(**/*.key)",
      "Read(**/.env.*)",
      "Bash(sudo:*)",
      "Bash(git reset --hard*)",
      "Bash(git reset --hard:*)",
      "Read(.env.*)"
    ],
    "ask": [
      "Bash(git push:*)",
      "Bash(git push *)",
      "Bash(git commit:*)",
      "Bash(rm *)",
      "Bash(rm:*)",
      "Bash(git rebase *)",
      "Bash(wget *)",
      "Bash(npm publish:*)",
      "Bash(git commit *)",
      "Bash(gh pr merge *)"
    ]
  }
}

Subagents (6)

codex-reviewer
model: claude-sonnet-5-5
Cross-vendor code review. The local OpenAI Codex CLI reviews a finished change set (uncommitted diff, branch, or commit) to catch what same-family Claude review misses. Separate small usage pool; one
codex
model: claude-haiku-5-5
Hands a self-contained, well-specified coding task to the local OpenAI Codex CLI (separate but small usage pool, use sparingly). For isolated work like one-shot scripts, boilerplate, or a contained bu
implementer
model: claude-sonnet-5-5
Mid-tier worker that implements a well-specified coding task: edits, new files, refactors, tests. Use whenever the plan is already decided and the remaining work is writing the code.
researcher
model: claude-haiku-5-5
Cheap read-only worker for external knowledge: API docs, release notes, changelogs, library READMEs, GitHub issues, and general web lookups. Use for "how does API X work / what changed in Y / is there
runner
model: claude-haiku-5-5
Cheap worker that runs builds, tests, and linters (xcodebuild, swift test, swiftlint, npm test, etc.) and returns only what failed. Use instead of running noisy build or test commands in the main loop
scout
model: claude-haiku-5-5
Cheap read-only worker for searching, locating code, reading files, and summarizing. Use for any "find where X is / what does Y look like" question instead of reading files in the main loop.

Similar rigs

copied ✓