bodencrouch/codex-better-subagents
Route Codex and Claude Code subagents to the right model and reasoning level, with ready-made worker definitions
stack: Shell
ARCHETYPE
Pragmatist
A balanced, no-drama setup: some rules, some tools, nothing extreme.
Copy this rig
# review before running: this installs third-party code $ npx degit bodencrouch/codex-better-subagents/agents ./rig-codex-better-subagents/agents $ npx degit bodencrouch/codex-better-subagents/skills ./rig-codex-better-subagents/skills
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.
$ npx degit bodencrouch/codex-better-subagents/skills/codex-better-subagents .claude/skills/codex-better-subagents $ curl -fsSL --create-dirs -o .claude/agents/basic-worker.md https://raw.githubusercontent.com/bodencrouch/codex-better-subagents/main/agents/claude/basic-worker.md $ curl -fsSL --create-dirs -o .claude/agents/deep-worker.md https://raw.githubusercontent.com/bodencrouch/codex-better-subagents/main/agents/claude/deep-worker.md $ curl -fsSL --create-dirs -o .claude/agents/frontier-worker.md https://raw.githubusercontent.com/bodencrouch/codex-better-subagents/main/agents/claude/frontier-worker.md $ curl -fsSL --create-dirs -o .claude/agents/standard-worker.md https://raw.githubusercontent.com/bodencrouch/codex-better-subagents/main/agents/claude/standard-worker.md
This rig commits no guardrails. Here is the community baseline instead — the deny/ask rules most often found across all 16.2k rigs:
{
"permissions": {
"deny": [
"Read(./.env)",
"Read(./.env.*)",
"Read(**/.env)",
"Read(~/.ssh/**)",
"Bash(git push --force:*)",
"Read(**/*.pem)",
"Read(.env)",
"Read(~/.aws/**)",
"Read(**/*.key)",
"Read(**/.env.*)",
"Bash(git push --force*)",
"Bash(rm -rf:*)",
"Bash(rm -rf *)",
"Bash(sudo *)",
"Bash(sudo:*)",
"Bash(rm -rf /)",
"Bash(git push -f:*)",
"Read(.env.*)",
"Bash(git reset --hard*)",
"Bash(rm -rf /*)"
],
"ask": [
"Bash(git push:*)",
"Bash(git push *)",
"Bash(gh pr merge *)",
"Bash(git commit:*)",
"Bash(rm *)",
"Bash(gh pr merge:*)",
"Bash(gh pr create *)",
"Bash(git reset --hard:*)",
"Bash(npm publish:*)",
"Bash(git rebase *)"
]
}
} Skills (1)
Subagents (4)
| basic-worker model: gpt-6-luna | Basic-tier worker on GPT-6 Luna at high effort. Use proactively for lookups, read-and-report, grep or inventory audits, extraction, classification, formatting, renames, summaries, small known edits, a |
| deep-worker model: claude-opus-5-5 | Deep-tier worker on Claude Opus 5.5 at medium effort. Use for ambiguous bugs, cross-file debugging, migrations, architecture, code review where a miss is costly, reverse-engineering hypotheses and par |
| frontier-worker model: gpt-6-astra | Frontier-tier worker on GPT-6 Astra at medium effort. Use for operating browsers, desktop apps or GUIs, document-heavy, scientific or security work, the hardest calls, and second opinions on deep-tier |
| standard-worker model: gpt-6.1-sol | Standard-tier worker on GPT-6.1 Sol at medium effort. Use for clear features, bounded bug fixes, refactors, docs, normal UI work, tests from a spec, and propagating confirmed facts. |
Similar rigs
BabuNfr666/claude-effort-router
Let Claude Code pick the reasoning effort for each task - a plugin that routes hard work to high/max-effort subagents.
Pragmatist 325 tok · · Q0
Ancienttwo/repo-harness
File-backed workflow harness for reliable Claude Code and Codex sessions.
Orchestrator 4.0k tok · · Q8
aimpact-inc/repo-harness
File-backed workflow harness for reliable Claude Code and Codex sessions.
Orchestrator 6.6k tok · · Q10
anomalyco/opencode
The open source coding agent.
Pragmatist 14.3k tok · · Q8