what-would-chad-do

name	what-would-chad-do
description	Run a final Chad reflection before closing implementation work. Use when coding work appears complete, momentum is slowing, or the agent is about to stop and needs to decide whether there is one more high-leverage, local, reversible step toward the real goal.
policy_doc_kind	skill
classification	canonical
canonical_owner	self
authority_level	procedural
in_verifier_scope	true

What Would Chad Do

Use this skill at the end of implementation work, not at the start.

The purpose is to prevent premature closure when there is still a clear, bounded, high-leverage next step.

Reflection

Before stopping, ask:

1. What is the real goal behind the task, not just the most recent subtask?
2. Is there one more small, local, reversible step that would materially improve the result now?
3. Would Chad keep going because the next move is obvious, safe, and aligned with the larger objective?

Continue If

• there is an obvious adjacent improvement that materially strengthens the outcome
• the current result works, but the path to the stated goal is still incomplete
• there is a missing verification, integration, polish, or follow-through step that is clearly in scope
• the autonomous-builder goal is still best served by one more bounded action

Stop If

• the user's actual goal is satisfied end to end
• evidence is complete and verification is already strong
• the remaining ideas are speculative, broad, or would open a new track of work
• continuing would require asking a real preference question or crossing an authority boundary

Bias

• prefer one more concrete step over a progress report
• prefer strengthening the system over narrating what was already done
• do not invent bonus projects
• do not loop forever; take the next best bounded step or stop with evidence