A curated collection of installable skills and sub-agents for Claude Code, covering deep research, image generation, GitHub automation, and more, with easy plugin marketplace installation.

A Claude Code plugin that provides a universal radial-tree exploration engine with swappable presets for divergent ideation, adversarial critique, design-space exploration, and code audit.

Autonomously improve a real artifact (code, training recipe, agent harness, data pipeline, prompt) against an objective and an evaluator, using Hypothesis Tree Refinement (HTR) from the Arbor paper. Use this whenever someone wants to iteratively optimize something over many experiments without overfitting — e.g. "get my model's eval score up", "improve this agent/harness", "tune this pipeline", "beat the baseline on this benchmark", "run a search over approaches and keep the best", "do an MLE-bench / Kaggle-style optimization", or any long-horizon "make this artifact better and don't just memorize the dev set" task. Trigger it even when the user doesn't say "Arbor" or "hypothesis tree" but describes repeated experiment-and-evaluate loops, branching exploration of competing ideas, or worries about a dev/test gap. Runs Claude itself as the coordinator with subagent executors in isolated git worktrees; for the standalone `arbor` CLI tool see references/arbor-upstream.md.

This list of writing prompts covers a range of topics and tasks, including brainstorming research ideas, improving language and style, conducting literature reviews, and developing research plans.

A spec-driven development and code quality toolkit for AI coding agents, installable via npm and integrated with multiple agent harnesses.

Claude Code skills: codex-sprint, review-loop (iterative simplify-review-fix), rust-dev (FAIL FAST standards)

Collection of 17 focused AI agent skills for tasks like brainstorming, research, and planning, compatible with skills.sh and installable via agent runtimes.

This repository hosts a suite of specialized agents designed to power your brainstorming sessions. Each agent brings a unique perspective and tailored approach to generate creative, targeted solutions for diverse challenges. Choose the right agent for your specific brainstorming needs and unlock innovative ideas with precision.

Parallel divergent ideation — spawns N isolated generator agents under different cognitive frames (regulator, biology, speedrunner, 10-year-old, zero-budget), then a critic pass scores, clusters, prunes traps, and deepens the top 3. Use for open-ended design, architecture, naming, API/SDK surface, and fuzzy debugging where the obvious answer is expensive to get wrong. Triggers "/adhd", "adhd mode", "brainstorm", "ideate", "widen the option space", "divergent ideas", "we keep landing on the same idea". Skip for lookups, syntax, bugs with a known root cause, or closed phrasing ("quick", "standard", "canonical", "textbook"). Use /oracle compare to evaluate options you already have — adhd generates the option space; use /plan-ceo-review to challenge whether to build at all.

Help a human author write an implementation-ready GitHub issue that Vigilante can execute reliably.

사용자가 /grill-me 를 호출하거나 아이디어·기능 요청·스펙을 구현 전에 굽고(grill)·압박 검증하고·심문해 달라고 할 때 사용. superpowers:brainstorming 의 opt-in 대체재 — 작업마다 사용자가 둘 중 하나를 고른다.

Pressure-test an existing plan or design artifact until it is coherent enough for the next workflow step or plainly blocked. Use when the user wants to be grilled on a plan, design doc, PRD, ADR, rollout plan, migration runbook, rolling-wave project shape, or slice contract; when they say "grill me"; or when Codex should interrogate a source-of-truth artifact, inspect code and adjacent docs before asking, auto-answer evidence-backed questions, ask only about conflicts or low-confidence recommendations, edit the original file in place, preserve the artifact's existing format, and keep conclusions, blockers, readiness, and deferred work synchronized in the document.