brackenjoe/agent-infrastructure
A guide for AI agent set up. A manager, a cheaper builder for code, any agents for other repetitive tasks, and the governance to keep them in line.
stack: Shell
ARCHETYPE
Minimalist
Lean instructions, few tools, tiny context tax. Lets the model think.
Copy this rig
# review before running: this installs third-party code
$ npx degit brackenjoe/agent-infrastructure/.claude ./rig-agent-infrastructure # inspect, then merge into .claude/ 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/builder.md https://raw.githubusercontent.com/brackenjoe/agent-infrastructure/main/starter/.claude/agents/builder.md $ curl -fsSL --create-dirs -o .claude/agents/manager.md https://raw.githubusercontent.com/brackenjoe/agent-infrastructure/main/starter/.claude/agents/manager.md $ curl -fsSL --create-dirs -o .claude/agents/reviewer.md https://raw.githubusercontent.com/brackenjoe/agent-infrastructure/main/starter/.claude/agents/reviewer.md
Merge into .claude/settings.json (project) or ~/.claude/settings.json (user). Hook commands reference scripts in the source repo — copy those too.
{
"permissions": {
"ask": [
"Edit(**/.claude/agents/**)"
]
}
} Subagents (3)
| builder model: sonnet | Builder for a bounded change to a settled design. The brief names the files it may edit, the data shape, the pattern to copy and the checks; one deliverable per brief. Not for design or anything live. |
| manager model: inherit | The session's default. Reviews each request, routes it to the agent whose description fits, or does it itself when none fits better; proposes a new agent when an area keeps coming back. |
| reviewer model: opus | Read-only review by an agent with none of the drafting context - text before it leaves the repository, a rule change, a pull-request or issue draft, a check script against its threat model. Returns ra |
Permissions
deny (0)
—
ask (1)
Edit(**/.claude/agents/**)
allow (0)
—
Similar rigs
Gildaciolopes/agentic-dev-stack
Smart vibe coding with Claude Code: multi-agent orchestration (Opus plans; Sonnet, Haiku, Gemini, Codex and free models execute), Spec Kit SDD, caveman + RTK + graphify token savings, sandboxed secrets, self-healing free models and a Hermes
Minimalist 1.2k tok · · Q10
CrepuscularIRIS/autopilot-dev
Autopilot Dev · 无人值守开发模板:扔一句需求,自动调研补全、TDD 开发、跨模型验收
Minimalist 1.2k tok · · Q0
aniteshdharam/claude-token-policy
A Claude Code plugin that cuts your token bill by shrinking context per call, not by picking a cheaper model: a policy skill, enforcing hooks, and a census script that measures your own transcripts.
Minimalist 247 tok · · Q0
hamitsarmis/strategy-team
Claude Code agent team that researches, backtests and vets trading strategies against pre-registered criteria. Engine, registry and Docker stack are built; it places no orders.
Pragmatist 855 tok · · Q15