using-superpowers

Enforces skill discovery and invocation before any agent response or action.

10|1|Updated Jul 29, 2026
One-click install
npx skills add https://github.com/flowing-abyss/obsidian-local-fonts --skill using-superpowers-flowing-abyss
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: using-superpowers
Source: https://github.com/flowing-abyss/obsidian-local-fonts/tree/main/.ai/skills/using-superpowers
Command: npx skills add https://github.com/flowing-abyss/obsidian-local-fonts --skill using-superpowers-flowing-abyss

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? AI agents often skip available skills and improvise answers, leading to inconsistent workflows and missed process steps. This Skill establishes a mandatory rule that relevant skills must be located and invoked before any response, clarifying question, or file exploration. ## Core Features & Use Cases - Mandatory Skill Invocation: Requires checking for and invoking applicable skills before any action, with a red-flags table that catches common rationalizations for skipping them. - Skill Priority Ordering: Defines how process skills (brainstorming, systematic-debugging) run before implementation skills, and where six vendored template skills slot into that ordering. - Cross-Platform Tool Mapping: Ships reference files mapping generic skill actions (dispatch subagent, create todo, read file) to concrete tools on Claude Code, Codex, Copilot CLI, Gemini CLI, Antigravity, and Pi. - Use Case: An agent starting a session in a multi-runtime repository loads this Skill first, announces which skill applies to the user's request, and follows that skill's checklist instead of improvising. ## Quick Start At the start of any conversation, invoke the relevant skill before responding and announce which skill you are using and why.

Frequently Asked Questions about using-superpowers

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I make an AI agent use skills before responding?

Load a skill-enforcement instruction at session start that requires checking for applicable skills before any response, including clarifying questions. This Skill defines that rule plus a red-flags table listing rationalizations to reject, such as treating simple questions as exempt.

What is the correct order for invoking process and implementation skills?

Process skills like brainstorming and systematic-debugging run first because they set the approach, then implementation skills carry out the work. For example, a build request triggers brainstorming before any coding skill, and a bug report triggers systematic-debugging first.

Does this skill work with Claude Code, Codex, and Gemini CLI?

Yes, it includes reference files mapping generic skill actions to each platform's concrete tools. Claude Code uses Read, Write, and Agent; Codex uses spawn_agent with multi_agent enabled; Gemini CLI uses invoke_agent and write_todos; Copilot CLI, Antigravity, and Pi have their own mappings.

How do subagents work across different AI CLI tools?

Subagent dispatch maps to Agent on Claude Code, spawn_agent on Codex, task on Copilot CLI, invoke_agent on Gemini CLI, and invoke_subagent on Antigravity. Pi has no built-in subagent tool, so the pi-subagents package or sequential execution is used instead.

When should a skill be skipped during a task?

A skill can be skipped only when it clearly does not apply after being invoked and reviewed, or when explicit user instructions override it. User instructions in files like CLAUDE.md or AGENTS.md take precedence over skills, which in turn override default behavior.