Adversarial multi-agent planning skill. Self-orchestrates 5 hostile category members (unspecified-low, unspecified-high, deep, ultrabrain, artistry) via team-mode for ruthless cross-critique debate, distills only the defensible insights, then MANDATORILY hands the distilled insight bundle to the `plan` agent for executable plan formalization.
安装
npx skills add code-yeongyu/oh-my-openagent --skill hyperplannpx skills add code-yeongyu/oh-my-openagent --skill hyperplan -a claude-code -gnpx skills add code-yeongyu/oh-my-openagent --skill hyperplan -a codex -ggemini skills install https://github.com/code-yeongyu/oh-my-openagent.git --path packages/shared-skills/skills/data-scientistgh skill install code-yeongyu/oh-my-openagent hyperplangit clone --depth 1 https://github.com/code-yeongyu/oh-my-openagent /tmp/code-yeongyu__oh-my-openagent
mkdir -p ~/.claude/skills ~/.agents/skills
cp -R /tmp/code-yeongyu__oh-my-openagent/packages/shared-skills/skills/data-scientist ~/.claude/skills/hyperplan # Claude Code
cp -R /tmp/code-yeongyu__oh-my-openagent/packages/shared-skills/skills/data-scientist ~/.agents/skills/hyperplan # Codex, Cursor, Gemini CLI, Copilot, OpenCode