AnirudhaRamesh/experiment-automation-claude-code-plugin
Claude Code setup for iterative experiment automation with notifications.
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 AnirudhaRamesh/experiment-automation-claude-code-plugin/agents ./rig-experiment-automation-claude-code-plugin/agents $ npx degit AnirudhaRamesh/experiment-automation-claude-code-plugin/skills ./rig-experiment-automation-claude-code-plugin/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 AnirudhaRamesh/experiment-automation-claude-code-plugin/skills/compare-configs .claude/skills/compare-configs $ npx degit AnirudhaRamesh/experiment-automation-claude-code-plugin/skills/experiment-loop-resume .claude/skills/experiment-loop-resume $ npx degit AnirudhaRamesh/experiment-automation-claude-code-plugin/skills/experiment-loop-status .claude/skills/experiment-loop-status $ npx degit AnirudhaRamesh/experiment-automation-claude-code-plugin/skills/experiment-loop .claude/skills/experiment-loop $ npx degit AnirudhaRamesh/experiment-automation-claude-code-plugin/skills/launch-and-monitor .claude/skills/launch-and-monitor $ npx degit AnirudhaRamesh/experiment-automation-claude-code-plugin/skills/monitor .claude/skills/monitor $ npx degit AnirudhaRamesh/experiment-automation-claude-code-plugin/skills/setup .claude/skills/setup $ npx degit AnirudhaRamesh/experiment-automation-claude-code-plugin/skills/team-lead .claude/skills/team-lead $ npx degit AnirudhaRamesh/experiment-automation-claude-code-plugin/skills/verify-experimental-setup .claude/skills/verify-experimental-setup
$ curl -fsSL --create-dirs -o .claude/agents/aichor-experiment-launcher.md https://raw.githubusercontent.com/AnirudhaRamesh/experiment-automation-claude-code-plugin/main/agents/aichor-experiment-launcher.md $ curl -fsSL --create-dirs -o .claude/agents/experiment-fixer.md https://raw.githubusercontent.com/AnirudhaRamesh/experiment-automation-claude-code-plugin/main/agents/experiment-fixer.md $ curl -fsSL --create-dirs -o .claude/agents/experiment-implementer.md https://raw.githubusercontent.com/AnirudhaRamesh/experiment-automation-claude-code-plugin/main/agents/experiment-implementer.md $ curl -fsSL --create-dirs -o .claude/agents/experiment-log-analyzer.md https://raw.githubusercontent.com/AnirudhaRamesh/experiment-automation-claude-code-plugin/main/agents/experiment-log-analyzer.md $ curl -fsSL --create-dirs -o .claude/agents/notion-tracker.md https://raw.githubusercontent.com/AnirudhaRamesh/experiment-automation-claude-code-plugin/main/agents/notion-tracker.md $ curl -fsSL --create-dirs -o .claude/agents/slack-notifier.md https://raw.githubusercontent.com/AnirudhaRamesh/experiment-automation-claude-code-plugin/main/agents/slack-notifier.md
This rig commits no guardrails. Here is the community baseline instead — the deny/ask rules most often found across all 6,974 rigs:
{
"permissions": {
"deny": [
"Read(./.env)",
"Read(**/.env)",
"Read(~/.ssh/**)",
"Bash(rm -rf *)",
"Read(**/*.pem)",
"Bash(rm -rf /)",
"Bash(git push --force:*)",
"Bash(sudo *)",
"Read(~/.aws/**)",
"Bash(rm -rf /*)",
"Read(./.env.*)",
"Read(.env)",
"Bash(git push --force*)",
"Bash(rm -rf:*)",
"Read(**/.env.*)",
"Read(**/*.key)",
"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(npm publish:*)",
"Bash(wget *)",
"Bash(git rebase *)",
"Bash(gh pr merge *)",
"Bash(git commit *)"
]
}
} Skills (9)
compare-configsexperiment_loop_resumeexperiment_loop_statusexperiment_looplaunch_and_monitormonitorsetupteam_leadverify-experimental-setup
Subagents (6)
| aichor-experiment-launcher model: sonnet | Use this agent when the user explicitly requests to launch a run/experiment on aichor, or when another agent has made code changes and needs to launch an experiment as directed by the user. Examples:\ |
| experiment-fixer model: opus | Use this agent to diagnose and fix failed AIChor experiments. Reads analysis reports and loop state, explores the codebase, and either applies a minimal fix or reports that the issue requires human in |
| experiment-implementer model: opus | Use this agent to implement an approved experiment plan into concrete code changes. Reads the plan from the experiment loop state file, explores the codebase, and makes targeted changes. Only invoked |
| experiment-log-analyzer model: sonnet | Use this agent when you need to analyze experiment logs, error traces, or execution failures to identify root causes and locate problematic code sections. Examples:\n\n<example>\nContext: A developer |
| notion-tracker model: sonnet | Use this agent to create and update Notion pages for experiment tracking. Handles: creating tracking pages in the Planning database, creating experiment log subpages, appending iteration entries, and |
| slack-notifier model: sonnet | Use this agent to send experiment status notifications to Slack. Reads state.json for context and sends formatted messages to configured channels. Supports concise and detailed message formats. |
Similar rigs
nizos/tdd-guard
Automated TDD enforcement for Claude Code
Minimalist 1.3k tok ·
terranc/claude-telegram-bot-bridge
A lightweight Telegram bot that bridges Claude Code to any local folder, with autostart support for remote mobile development.
Minimalist 2.1k tok ·
stefanoginella/aicontainer
Sandboxed devcontainer for running Claude Code, Codex, and OpenCode in bypass / auto-approve mode.
Pragmatist 13.0k tok ·
MrTig-afk/claude-remote
Claude Code on your PC, started from your phone, with your full setup intact.
Pragmatist 2.8k tok ·