Claude Skills by cukas

Register a new AI CLI as a buddy — interactive wizard or one-liner

Multi-AI confidence bid — each AI rates their confidence on a task, user picks who takes it

Open multi-AI conversation — all buddies think together on a topic, no competition, just talk

Code review via OpenAI Codex — review uncommitted changes, branches, or commits

Ask OpenAI Codex anything — brainstorm, delegate tasks, get a second opinion

Evolutionary multi-AI code forge — three AIs build, test, and cross-pollinate

Code review via Google Gemini — review uncommitted changes, branches, or commits

Ask Google Gemini anything — brainstorm, delegate tasks, get a second opinion

Show ELO ratings leaderboard — per task class or overall

Code review via OpenCode — review uncommitted changes, branches, or commits

Ask OpenCode anything — brainstorm, delegate tasks, get a second opinion

Multi-mode AI debate — adversarial, Socratic, steelman, red-team, synthesis, or postmortem with evidence

Kill current session and seamlessly transfer context to a new one. One click. Zero planning.

Save a structured handoff document so another Claude instance (or this one after /clear) can resume the work. Use when context is getting long, switching tasks, or before /clear.

Guided pruning of stale memories with user confirmation

Browse and search Pensieve session memories — what files were touched, commands run, errors resolved across sessions. Use to understand session history when context is unclear.

Resume work from a previous session with state verification and git patch restore. Reads the handoff, verifies files and git state, restores patches, and presents a structured briefing.

Manage Time-Turner parallel agents — check status, diff changes, merge results, or cancel. Time-Turner spawns a headless Claude agent to continue work when context runs low.

Use when implementing features, refactors, or multi-file changes. Orchestrated plan-then-execute with subagent dispatch and regression checkpoints.

Use when investigating a bug or unexpected behavior. Forces structured root-cause investigation before any code changes. Trace the call chain, document hypotheses, confirm the root cause, then fix.

Use after completing any code implementation, bug fix, or refactor. Structured review with pass/fail verification before declaring done.

Use when fixing a bug, unexpected behavior, or test failure. Investigation-first TDD loop that traces root cause before touching code.