<SUBAGENT-STOP> If you were dispatched as a subagent to execute a specific task, skip this skill. </SUBAGENT-STOP> <EXTREMELY-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. </EXTREMELY-IMPORTANT>

Instruction Priority

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

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

If CLAUDE.md or 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 to Access Skills

In Claude Code: Use the Skill tool. When you invoke a skill, its content is loaded and presented to you—follow it directly. Never use the Read tool on skill files.

In Cursor / OpenCode / Codex CLI: Check your platform's documentation for how skills are loaded. Cross-platform config files are in .cursor/, .opencode/, and .codex/ respectively.

Platform Adaptation

Skills use Claude Code tool names. Non-CC platforms: see references/codex-tools.md (Codex) for tool equivalents.

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];
    "About to EnterPlanMode?" [shape=doublecircle];
    "Already brainstormed?" [shape=diamond];
    "Invoke brainstorming skill" [shape=box];
    "Might any skill apply?" [shape=diamond];
    "Invoke Skill tool" [shape=box];
    "Announce: 'Using [skill] to [purpose]'" [shape=box];
    "Has checklist?" [shape=diamond];
    "Create TodoWrite todo per item" [shape=box];
    "Follow skill exactly" [shape=box];
    "Respond (including clarifications)" [shape=doublecircle];

    "About to EnterPlanMode?" -> "Already brainstormed?";
    "Already brainstormed?" -> "Invoke brainstorming skill" [label="no"];
    "Already brainstormed?" -> "Might any skill apply?" [label="yes"];
    "Invoke brainstorming skill" -> "Might any skill apply?";

    "User message received" -> "Might any skill apply?";
    "Might any skill apply?" -> "Invoke Skill tool" [label="yes, even 1%"];
    "Might any skill apply?" -> "Respond (including clarifications)" [label="definitely not"];
    "Invoke Skill tool" -> "Announce: 'Using [skill] to [purpose]'";
    "Announce: 'Using [skill] to [purpose]'" -> "Has checklist?";
    "Has checklist?" -> "Create TodoWrite todo per item" [label="yes"];
    "Has checklist?" -> "Follow skill exactly" [label="no"];
    "Create TodoWrite todo per item" -> "Follow skill exactly";
}

Red Flags

These thoughts mean STOP—you're rationalizing:

Thought	Reality
"This is just a simple question"	Questions are tasks. Check for skills.
"I need more context first"	Skill check comes BEFORE clarifying questions.
"Let me explore the codebase first"	Skills tell you HOW to explore. Check first.
"I can check git/files quickly"	Files lack conversation context. Check for skills.
"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.
"This doesn't count as a task"	Action = task. Check for skills.
"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.

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 (api-design, deployment-patterns) - these guide execution

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

Skill Types

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

Flexible (patterns): Adapt principles to context.

The skill itself tells you which.

User Instructions

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

Two-Layer Agent Architecture

Superpowers-ecc uses two complementary layers for agent-based work:

Layer 1: Embedded Subagent Prompts

Skills like subagent-driven-development contain execution machinery — they define how to decompose work, dispatch subagents via the Task tool, and coordinate results. These are process skills that drive multi-agent workflows from within a conversation.

When to use: You're executing a multi-step implementation plan and need to parallelize work across subagents. The skill itself orchestrates the dispatch.

Key skills: subagent-driven-development, dispatching-parallel-agents, executing-plans

Layer 2: Standalone Agents

The agents/ directory contains domain-expert agents — specialized consultants you dispatch via commands. Each agent has a focused system prompt and a narrow domain of expertise.

Agent	Domain
planner	Breaks down tasks into phased implementation plans
architect	System design, component boundaries, dependency analysis
code-reviewer	Code quality, patterns, correctness
security-reviewer	OWASP-based security analysis
build-error-resolver	Diagnoses and fixes build/compile errors
refactor-cleaner	Identifies and executes safe refactors
e2e-runner	Runs and debugs end-to-end test suites

When to use: You need a focused expert opinion or to delegate a well-scoped task (review this PR, analyze this architecture, fix this build error). Dispatch via the appropriate command.

Choosing Between Layers

Situation	Use
Multi-step implementation with parallelism	Layer 1 (embedded subagent prompts)
Focused expert task (review, analysis, diagnosis)	Layer 2 (standalone agents)
Complex project requiring both coordination and expertise	Both — Layer 1 orchestrates, Layer 2 agents get dispatched as subagents

State & Learning System

Superpowers-ecc persists state across conversations in ~/.claude/superpowers-ecc/.

Sessions

Sessions maintain context across multiple conversations on the same project or task. When a session is active, the agent has access to prior decisions, plans, and progress — avoiding the cold-start problem of each conversation starting from scratch.

Instincts

The instinct system accumulates learned patterns with confidence scores. As the agent encounters repeated situations, successful approaches get reinforced and unsuccessful ones get deprioritized. Instincts represent "muscle memory" — things the agent has learned work well in this codebase or for this user.

Each instinct has:

• A pattern describing the situation
• A response describing the learned behavior
• A confidence score (0.0–1.0) that increases with successful application

Commands

Command	Purpose
/learn	Capture a new learning from the current context
/evolve	Review and update existing instincts based on new evidence
/instinct-status	Show current instincts and their confidence scores

ECC Additions

Beyond the 14 superpowers-origin skills, superpowers-ecc adds 10 skills that extend coverage into security, operations, continuous learning, and development rigor:

Skill	What it adds
security-review	OWASP-based security checklist for code and architecture review
continuous-learning-v2	Instinct system with confidence scoring — the learning engine behind /learn, /evolve, /instinct-status
search-first	Enforces research-before-coding discipline — read docs, search issues, check prior art
eval-harness	Eval-driven development with grader types (exact, llm, human) and pass@k metrics
verification-loop	Continuous build/test/lint/typecheck verification — catches breakage as you code
strategic-compact	Token-aware compaction suggestions — when to compact and what to preserve
token-optimization	System prompt slimming, model selection guidance, context budget management
api-design	REST patterns, pagination, error responses, versioning conventions
deployment-patterns	CI/CD pipelines, Docker configuration, health checks, rollback strategies
e2e-testing	Playwright end-to-end testing patterns, selectors, fixtures, debugging

Skill Inventory

All 24 skills grouped by category:

Core Workflow (7)

Skill	Purpose
brainstorming	Structured ideation before implementation
writing-plans	Phased implementation plans with clear deliverables
executing-plans	Plan execution with progress tracking
test-driven-development	Red-green-refactor TDD discipline
systematic-debugging	Hypothesis-driven bug investigation
subagent-driven-development	Multi-agent parallel execution
dispatching-parallel-agents	Parallel subagent dispatch patterns

Supporting (6)

Skill	Purpose
finishing-a-development-branch	Branch completion checklist
requesting-code-review	How to request and structure code reviews
receiving-code-review	How to process and act on review feedback
verification-before-completion	Final verification before marking work done
using-git-worktrees	Git worktree workflows for parallel development
writing-skills	How to author new skills

Onboarding (1)

Skill	Purpose
using-superpowers-ecc	This skill — how to find and use all other skills

Security & Quality (2)

Skill	Purpose
security-review	OWASP-based security analysis
verification-loop	Continuous build/test/lint/typecheck verification

Development Rigor (3)

Skill	Purpose
search-first	Research before coding
eval-harness	Eval-driven development with grader types and pass@k
e2e-testing	Playwright end-to-end testing patterns

Architecture & API (2)

Skill	Purpose
api-design	REST patterns, pagination, error responses
deployment-patterns	CI/CD, Docker, health checks

Learning & Optimization (3)

Skill	Purpose
continuous-learning-v2	Instinct system with confidence scoring
strategic-compact	Token-aware compaction suggestions
token-optimization	System prompt slimming, model selection