using-superpowers

name	using-superpowers
description	Use when implementing new features or applications, or starting complex multi-step tasks that may benefit from structured workflows like brainstorming, TDD, or debugging. NOT for simple questions or straightforward operations.

Subagents: If you were dispatched as a subagent to execute a specific task, skip this skill.

IMPORTANT: If you think there is even a 1% chance a skill might apply to what you are doing, you ABSOLUTELY MUST invoke the skill.

IF A SKILL APPLIES TO YOUR TASK, YOU DO NOT HAVE A CHOICE. YOU MUST USE IT.

This is not negotiable. This is not optional. You cannot rationalize your way out of this.

Simple Operations — No Skill Needed

DO NOT invoke skills or dispatch sub-agents for operations that are trivially fast and carry no risk of errors:

• Reading 1-3 files to understand code, configuration, or output
• Making a single edit to a file
• Running a simple command (ls, grep, find, git status, etc.)
• Answering a question based on information you already have
• Searching the codebase for a string or pattern
• Inspecting test output or error logs

These are tool calls, not tasks. Use them directly.

Sub-agents are for work that is:

• Multi-step: Requires 3+ distinct actions with judgment between them
• Substantive: Involves implementation, debugging, or design decisions
• Risk-bearing: Could introduce bugs if done incorrectly
• Time-consuming: Would take more than a handful of tool calls to complete

A sub-agent adds overhead (context construction, dispatch, result parsing). Only dispatch when the work justifies that cost.

Instruction Priority

Superpowers skills override default system prompt behavior, but user instructions always take precedence:

1. User's explicit instructions (AGENTS.md, direct requests) — highest priority
2. Superpowers skills — override default system behavior where they conflict
3. Default system prompt — lowest priority

If AGENTS.md says "don't use TDD" and a skill says "always use TDD," follow the user's instructions. The user is in control.

How Skills Work

Skills are auto-discovered and listed in your system prompt. When a skill applies, read its SKILL.md file and follow its instructions. Subagent dispatch is handled by the Task tool (see references/pi-tools.md).

Using Skills

The Rule

Invoke relevant or requested skills BEFORE any response or action. Even a 1% chance a skill might apply means that you should invoke the skill to check. If an invoked skill turns out to be wrong for the situation, you don't need to use it.

digraph skill_flow {
    "User message received" [shape=doublecircle];
    "Building something new?" [shape=doublecircle];
    "Proposal + feature spec exist?" [shape=diamond];
    "Invoke brainstorming skill" [shape=box];
    "Simple operation?" [shape=diamond];
    "Might any skill apply?" [shape=diamond];
    "Read SKILL.md" [shape=box];
    "Announce: 'Using [skill] to [purpose]'" [shape=box];
    "Has checklist?" [shape=diamond];
    "Create task tracking per item" [shape=box];
    "Follow skill exactly" [shape=box];
    "Do it directly" [shape=box, style=filled, fillcolor=lightgreen];
    "Respond (including clarifications)" [shape=doublecircle];

    "Building something new?" -> "Proposal + feature spec exist?";
    "Proposal + feature spec exist?" -> "Invoke brainstorming skill" [label="no — artifacts missing"];
    "Proposal + feature spec exist?" -> "Simple operation?" [label="yes — artifacts exist"];
    "Invoke brainstorming skill" -> "Simple operation?";

    "User message received" -> "Simple operation?";
    "Simple operation?" -> "Do it directly" [label="yes"];
    "Simple operation?" -> "Might any skill apply?" [label="no"];
    "Might any skill apply?" -> "Read SKILL.md" [label="yes, even 1%"];
    "Might any skill apply?" -> "Respond (including clarifications)" [label="definitely not"];
    "Read SKILL.md" -> "Announce: 'Using [skill] to [purpose]'";
    "Announce: 'Using [skill] to [purpose]'" -> "Has checklist?";
    "Has checklist?" -> "Create task tracking per item" [label="yes"];
    "Has checklist?" -> "Follow skill exactly" [label="no"];
    "Create task tracking per item" -> "Follow skill exactly";
}

Red Flags

These thoughts mean STOP—you're rationalizing:

Thought	Reality
"I need more context first"	Skill check comes BEFORE clarifying questions.
"Let me gather information first"	Skills tell you HOW to gather information.
"This doesn't need a formal skill"	If a skill exists, use it.
"I remember this skill"	Skills evolve. Read current version.
"The skill is overkill"	Simple things become complex. Use it.
"I'll just do this one thing first"	Check BEFORE doing anything.
"This feels productive"	Undisciplined action wastes time. Skills prevent this.
"I know what that means"	Knowing the concept ≠ using the skill. Invoke it.

Exception: The "Simple Operations" section above overrides these red flags. Reading files, making small edits, or running quick commands is NOT rationalization — it's the correct behavior. Only invoke skills when the work is substantive enough to warrant a structured process.

Skill Priority

When multiple skills could apply, use this order:

1. Process skills first (brainstorming, debugging) - these determine HOW to approach the task
2. Implementation skills second (domain-specific) - these guide execution

"Let's build X" → brainstorming first, then implementation skills. "Fix this bug" → debugging first, then domain-specific skills.

What counts as "already brainstormed": The brainstorming skill has been completed when the proposal (docs/design/YYYY-MM-DD-<topic>-proposal.md) and the feature spec (docs/design/YYYY-MM-DD-<topic>-spec.md) both exist. A conversation about the idea is not the same as brainstorming. If the artifacts don't exist, invoke the brainstorming skill — even if you've already discussed the idea at length.

Skill Types

Rigid (TDD, debugging): Follow exactly. Don't adapt away discipline.

Flexible (patterns): Adapt principles to context.

The skill itself tells you which.

Instruction Precedence

Instructions say WHAT, not HOW. "Add X" or "Fix Y" doesn't mean skip workflows.