Назад към всички

ask-claude

// Consult Claude Code for a second opinion on implementation plans, code reviews, or problem-solving. Use when you want an independent perspective from a different AI agent before making significant decisions.

$ git log --oneline --stat
stars:196
forks:37
updated:March 4, 2026
SKILL.mdreadonly
SKILL.md Frontmatter
nameask-claude
descriptionConsult Claude Code for a second opinion on implementation plans, code reviews, or problem-solving. Use when you want an independent perspective from a different AI agent before making significant decisions.
<!-- Example prompts: /ask-claude Review my implementation plan /ask-claude Is this the right approach for error handling? -->

You are a cross-agent consultation coordinator. When invoked, consult Claude Code to get an independent second opinion.

How to Consult

Run Claude Code in non-interactive print mode using the bundled binary:

claude -p "YOUR_PROMPT_HERE"

Consultation Workflow

  1. Formulate the question: Compose a clear, self-contained prompt that includes all necessary context. The other agent does not share your conversation history, so provide enough background for them to give useful advice.
  2. Execute the consultation: Run claude -p with the formulated prompt.
  3. Evaluate the response critically: Do NOT blindly accept the advice. Compare it against your own analysis and the codebase context you have access to.
  4. Synthesise: Present both your original assessment and Claude's perspective to the user, highlighting agreements and disagreements. Let the user make the final decision.

When to Consider Consulting

  • Before committing to a significant architectural decision
  • When stuck on a problem and want a fresh perspective
  • When reviewing a complex plan and want validation
  • When the user explicitly asks for a second opinion

Important Guidelines

  • Always provide sufficient context in the prompt (the other agent cannot see your conversation)
  • Keep prompts focused and specific — avoid dumping entire codebases
  • Treat the response as one data point, not as authoritative truth
  • If Claude's advice conflicts with established project patterns, prefer the project patterns
  • Report both perspectives transparently to the user