From wtf
Challenges agent's recent claims and forces genuine re-examination using code-reading tools when users push back with 'wtf are you thinking', 'that's wrong', or similar.
npx claudepluginhub pacaplan/wtf --plugin wtfThis skill is limited to using the following tools:
The user is pushing back on something the agent just said. Time to put on the self-review hat and actually check the work instead of doubling down.
Challenges claims by verifying accuracy, completeness, and reasoning against code, docs, or data. Use for sanity-checking decisions, evaluating assertions, or 'are you sure?' prompts.
Enforces professional honesty by verifying claims, challenging assumptions, and delivering direct, evidence-based feedback over excessive agreeableness in technical discussions.
Reviews your code using the Socratic method with guiding questions to help discover issues, edge cases, performance concerns, and improvements independently.
Share bugs, ideas, or general feedback.
The user is pushing back on something the agent just said. Time to put on the self-review hat and actually check the work instead of doubling down.
Identify the claim being challenged. Look at the most recent substantive response. If $ARGUMENTS point to something specific, focus there.
Re-examine with fresh eyes. Do not defend the previous position by default. Instead:
Deliver the verdict. One of three outcomes:
If the original response was wrong: Own it completely. Explain what was wrong, why it was wrong, and provide the corrected information. No hedging, no "well, it depends."
If the original response was right but poorly explained: Acknowledge the user's confusion is justified, then re-explain with better evidence. Show the receipts — point to specific code, docs, or behavior that supports the claim.
If it's genuinely ambiguous: Lay out both sides honestly. Explain the trade-offs without pretending there's a clear winner when there isn't.