Kuberwastaken

Prepare, validate, publish, and verify Megaphone releases. Use when the user asks to release Megaphone, choose or bump its version, update release notes, validate release readiness, create a stable vX.Y.Z tag, inspect the moving Dev release, publish Megaphone.dmg, verify updater compatibility, or confirm that megaphone.kuber.studio reflects the latest GitHub release.

Parallel autonomous ML research agents with a Director, git worktrees for per-agent experiment branches, a Skills library for validated technique reuse, a Synthesizer that distills collective knowledge overnight, and circadian rhythm (leisure 03:00–06:00 for paper reading and creative thinking). Uses OpenClaw sessions_spawn, cron, and steer natively. Use when: (1) start or run ML research agents overnight, (2) check agent status or experiment results, (3) view leaderboard or morning digest, (4) steer or stop agents, (5) ask what agents discovered or are exploring, (6) set up Litmus for the first time. NOT for: general coding, non-ML tasks, or machines without a GPU.

Install, set up, and use Reference — one MCP server that lets an AI agent search the past session transcripts AND memory files (CLAUDE.md, AGENTS.md, memory/*.md) of every AI coding tool on the machine (Claude Code, Codex CLI, Cursor, VS Code, …). Use this skill when the user wants cross-tool recall ("what did we decide in Codex?", "find the session where we wired X", "search my past chats"), when setting Reference up in a new host, or when an agent should remember work done in a different tool. Local-first, read-only, offline BM25 — nothing leaves the machine.

Set up and manage a personal cloud VPS as a dedicated remote workbench for AI coding agents (Claude Code + OpenAI Codex). Covers provisioning (git/node/pnpm/docker/tmux), installing & authenticating the agents, a self-aware agent environment, Tailscale networking (SSH with no exposed ports), a tmux + cmux cockpit, localhost port tunneling, reverse VPS→laptop file access (including a no-admin path for managed/work laptops), one-time migration of your skills/memory/history/credentials, and automatic two-way sync. Use this when the user wants to run AI agents on a server, offload heavy or parallel agent work off their laptop, "set up a VPS for Claude Code / Codex", spin up a remote agent box, or keep an agent setup synced across machines.

Spawn and control non-Claude coding-agent subagents (OpenAI Codex CLI and Cursor CLI — grok-4.5, composer-2.5, gpt-5.5) as one-prompt-in/one-answer-out subagents via the `subagent` CLI. Use when you want to offload a task to a Codex or Cursor subagent, get a second opinion from a different model family, run cross-model consensus, or fan one prompt out across several models in parallel.