clarify

Verify user intent and explain actual behavior with code and documentation evidence.

439|45|Updated Feb 17, 2026
One-click install
npx skills add https://github.com/umputun/cc-thingz --skill clarify-umputun
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: clarify
Source: https://github.com/umputun/cc-thingz/tree/main/plugins/workflow/skills/clarify
Command: npx skills add https://github.com/umputun/cc-thingz --skill clarify-umputun

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Handle user confusion by verifying intent, explaining actual behavior, and determining if there's a real issue to address.

Core Features & Use Cases

  • Identify confusion signals and categorize them (memory gap, project mixing, outdated mental model, documentation gaps, or configuration issues), then guide the user toward resolution.
  • Investigate by reading relevant code, checking configuration, and reviewing documentation to gather evidence and provide a grounded explanation.
  • If a real issue is detected, propose entering plan mode to design and implement a fix, including evidence-based reasoning and a clear next-step plan.

Quick Start

Describe the confusion and request an evidence-based explanation to determine if a real issue exists.

Frequently Asked Questions about clarify

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I debug code when I'm confused about why it behaves unexpectedly?

Debugging unexpected code behavior requires gathering evidence by reading relevant code, checking configuration, and reviewing documentation. This approach verifies your intent, explains actual behavior, and determines whether a real issue exists before proposing a fix.

What's the best way to analyze code and configuration to find the root cause of confusion?

Analyzing code and configuration to resolve confusion involves investigating the actual codebase, checking settings, and reviewing documentation. This evidence-based process categorizes the confusion source—such as memory gaps, project mixing, or outdated mental models—to provide a grounded explanation.

Can I get an evidence-based explanation for why my code isn't working as documented?

Evidence-based explanations for code not working as documented are generated by cross-referencing actual code behavior with available documentation. This process identifies whether the issue stems from documentation gaps, configuration issues, or a real bug in the code.

Does this approach help when I have an outdated mental model of how the project works?

Yes, it specifically identifies confusion signals like an outdated mental model or project mixing. By verifying your original intent against the actual codebase and configuration, it provides a grounded explanation that corrects your understanding of the project.

How do I move from code analysis to planning a fix when a real design flaw is detected?

When code analysis detects a real bug or design flaw, you can enter plan mode to design and implement a fix. This mode provides evidence-based reasoning and a clear, actionable next-step plan to resolve the identified issue.

What are the limitations of evidence-based code analysis for resolving confusion?

Evidence-based code analysis for resolving confusion is limited by the availability and accuracy of existing code, configuration, and documentation. Without accessible evidence, the tool cannot provide grounded explanations or determine if a real issue exists.