albiol2004/trio-agent-loop
Research-hardened two-agent coding loop (Lead + adversarial Evaluator) for Claude Code, with a portable driver for Codex, Cursor, OpenCode, Pi, Hermes, Gemini and more — agents communicate via markdown mailboxes
ARCHETYPE
Orchestrator
A bench of specialised subagents. The main agent mostly delegates.
Copy this rig
# review before running: this installs third-party code
$ npx degit albiol2004/trio-agent-loop/.claude ./rig-trio-agent-loop # 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.
$ npx degit albiol2004/trio-agent-loop/.agents/skills/trio-init .claude/skills/trio-init $ npx degit albiol2004/trio-agent-loop/.agents/skills/trio .claude/skills/trio $ npx degit albiol2004/trio-agent-loop/.claude/skills/trio-productionize .claude/skills/trio-productionize $ npx degit albiol2004/trio-agent-loop/.claude/skills/trio-ship .claude/skills/trio-ship $ npx degit albiol2004/trio-agent-loop/bridge/skills/trio-bridge .claude/skills/trio-bridge
$ curl -fsSL --create-dirs -o .claude/agents/trio-acceptance.md https://raw.githubusercontent.com/albiol2004/trio-agent-loop/main/.claude/agents/trio-acceptance.md $ curl -fsSL --create-dirs -o .claude/agents/trio-builder.md https://raw.githubusercontent.com/albiol2004/trio-agent-loop/main/.claude/agents/trio-builder.md $ curl -fsSL --create-dirs -o .claude/agents/trio-evaluator.md https://raw.githubusercontent.com/albiol2004/trio-agent-loop/main/.claude/agents/trio-evaluator.md $ curl -fsSL --create-dirs -o .claude/agents/trio-lead.md https://raw.githubusercontent.com/albiol2004/trio-agent-loop/main/.claude/agents/trio-lead.md $ curl -fsSL --create-dirs -o .claude/agents/trio-repair.md https://raw.githubusercontent.com/albiol2004/trio-agent-loop/main/.claude/agents/trio-repair.md $ curl -fsSL --create-dirs -o .claude/agents/trio-scout.md https://raw.githubusercontent.com/albiol2004/trio-agent-loop/main/.claude/agents/trio-scout.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 *)"
]
}
} Skills (5)
Subagents (6)
| trio-acceptance model: opus | Independent Trio acceptance author (r19) for the trio-native workflow. Turns GOAL.md into frozen black-box checks before any code is written; same tier as the Evaluator. Never the Lead, the Evaluator |
| trio-builder model: sonnet | Sonnet primary implementation worker for the trio loop. Executes one well-specified increment handed down by the Opus Lead, including substantive application logic, tests, and integration work. |
| trio-evaluator model: opus | Opus adversarial evaluator of the duo loop. Verifies the Lead's iteration against PLAN.md's acceptance criteria by actually exercising the code, using Sonnet explorers for scoped reconnaissance and So |
| trio-lead model: opus | Opus lead of the duo loop — plans, delegates the main implementation pass to Sonnet builders, then reviews and corrects their work. Maintains PLAN.md and REPORT.md and owns the final result. |
| trio-repair model: sonnet | Sonnet scoped-repair worker for the trio loop. Invoked on VERDICT: ITERATE scope=local:<paths> — fixes exactly the listed failure scope with no re-planning, refactoring, or scope expansion. |
| trio-scout model: sonnet | Sonnet read-only explorer for the trio loop. Answers codebase questions for the Opus Lead/Evaluator — how things work, where things live, what conventions exist. Never modifies anything. |
Similar rigs
rtk-ai/rtk
CLI proxy that reduces LLM token consumption by 60-90% on common dev commands. Single Rust binary, zero dependencies
Orchestrator 2.6k tok ·
Yeachan-Heo/oh-my-claudecode
Teams-first Multi-agent orchestration for Claude Code
Orchestrator 14.5k tok ·
alirezarezvani/claude-skills
380 Claude Code skills & agent skills & plugins (30+ Agents, 70+ custom commands, 380+ skills, customizable references, scripts)for Claude Code, Codex, Gemini CLI, Cursor, and 8 more coding agents — engineering, marketing, product, complian
Orchestrator 48.5k tok ·
czlonkowski/n8n-mcp
A MCP for Claude Desktop / Claude Code / Windsurf / Cursor to build n8n workflows for you
Orchestrator 3.8k tok ·