using-superpowers

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

Updated Mar 26, 2026
One-click install
npx skills add https://github.com/erwinv2k-TKG/AgentesVSC --skill using-superpowers-erwinv2k-tkg
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: using-superpowers
Source: https://github.com/erwinv2k-TKG/AgentesVSC/tree/main/packs/superpowers/skills/using-superpowers
Command: npx skills add https://github.com/erwinv2k-TKG/AgentesVSC --skill using-superpowers-erwinv2k-tkg

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? AI assistants often skip available skills and answer directly, missing established workflows. This Skill establishes a mandatory protocol that forces the assistant to check for and invoke relevant skills before responding, asking clarifying questions, or taking any action. ## Core Features & Use Cases - Mandatory Skill Check: Requires invoking any potentially relevant skill (even at 1% likelihood) before any response, with a decision flowchart and a red-flags table to catch rationalizations. - Instruction Priority Model: Defines a clear hierarchy where user instructions (CLAUDE.md, GEMINI.md, AGENTS.md) override skills, and skills override default system behavior. - Cross-Platform Tool Mapping: Ships reference guides mapping Claude Code tools to Codex and Gemini CLI equivalents, including subagent and task-tracking differences. - Use Case: When a user asks to fix a bug, the assistant first invokes the debugging skill, announces which skill it is using, creates todos from the skill checklist, and only then proceeds with the fix. ## Quick Start Start a new conversation and ask the assistant to fix a bug, then confirm it invokes the relevant skill before responding.

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 assistant always use skills before responding?

Load this skill at conversation start so it establishes the rule that any potentially relevant skill must be invoked before any response, including clarifying questions. It provides a decision flowchart and red-flags list to prevent the assistant from rationalizing skipping skills.

How do skills work in Gemini CLI versus Claude Code?

In Claude Code you invoke skills with the Skill tool, while Gemini CLI uses the activate_skill tool and loads skill metadata at session start. The included gemini-tools.md reference maps tools like Read to read_file and TodoWrite to write_todos.

Do user instructions override skill instructions?

Yes, user instructions always take precedence. The priority order is explicit user instructions in CLAUDE.md, GEMINI.md, or AGENTS.md first, then skill guidance, then default system prompt behavior.

Can I use Claude Code skills in Codex?

Yes, the codex-tools.md reference maps Claude Code tools to Codex equivalents, such as Task to spawn_agent and TodoWrite to update_plan. Subagent-based skills require enabling multi_agent in the Codex config.toml file.

What happens when multiple skills could apply to one task?

Process skills like brainstorming or debugging are invoked first because they determine how to approach the task, followed by implementation skills that guide execution. The skill also distinguishes rigid skills that must be followed exactly from flexible pattern skills.