auto-ralphProactively detects tasks suitable for Ralph Loop iteration. Auto-scores tasks, generates prompts, and executes with configurable iterations.
1.0.0
3
debateMulti-model adversarial debate. Claude defends positions against external CLI models (Gemini, Codex, Qwen). Requires at least 1 external CLI. Produces tradeoff documents exposing hidden assumptions. Commands: /debate, /debate:doctor, /debate:adr
1.0.0
4
diagnosticianulElite Senior Principal Engineer persona (Top 0.01%). Dr. House of IT: cynical code diagnostics with 4 protocols — critic (code autopsy), architect (system design), visual (UI dictatorship), core (algorithmic surgery).
2026.1.0
5
openclawULTRA deep OpenClaw expertise from all 264 documentation pages. Covers architecture, channels, providers, tools, CLI, multi-agent routing, security, automation, platforms, installation, and troubleshooting.
1.0.0
6
orchestratorMulti-agent orchestration with task decomposition, parallel execution, and smart agent routing. Commands: /orchestrate, /parallel, /plan-only
1.0.0
7
statuslinePremium powerline-style statusline for Claude Code with git status, context window, and model info