using-superpowers-ecc

Skill-first workflow enforcement requires agents to invoke the skill-discovery step before producing any response. This prevents undisciplined behavior by ensuring the agent finds and follows the correct workflow or specialist skill before answering.

Agents skip workflow steps due to a lack of process discipline and skill-first enforcement. Without mandating a skill discovery step before responding, the agent defaults to ad-hoc behavior instead of loading the correct debugging or brainstorming skill checklist.

Yes, you can map Claude Code tool names to equivalents for Codex, Cursor, and OpenCode. The skill-first enforcement policy includes platform adaptation rules that explain where to find and configure cross-platform equivalents.

Instruction precedence establishes that user prompts take the highest priority, followed by superpowers-ecc skills, and then the default system. This policy ensures user requirements override default agent behaviors while maintaining process discipline.

Dispatch standalone agents or embedded subagents based on explicit policy rules defined during skill discovery. The enforcement framework dictates invocation ordering and provides explicit guidance on when to use subagents for tasks like security review or verification.