Check Review Alignment

The check-review-alignment skill verifies that AI-generated research outputs strictly adhere to the original user instructions and review criteria. It acts as a quality gate, ensuring consistency between initial prompts, iterative feedback, and final deliverables in complex workflows like Chinese academic writing.

When to use it

• Validating that a generated LaTeX paper matches specific formatting rules requested in the prompt.
• Ensuring an AI agent's response to a reviewer's critique does not drift from the core research topic.
• Confirming that multi-step instructions for data analysis are followed exactly in the final report.
• Checking if a translated document retains all constraints and nuances from the source language instructions.

Key capabilities

• Cross-references current output against stored initial instructions.
• Identifies deviations between review feedback and the agent's subsequent actions.
• Validates adherence to specific structural or stylistic guidelines provided by the user.

Example prompts

• "Run check-review-alignment on this draft to ensure it follows my original constraints for a Chinese literature review."
• "Use check-review-alignment to verify that the agent's response to the peer review comments didn't introduce new, unapproved methodologies."
• "Execute check-review-alignment to confirm the final LaTeX output matches the sectioning requirements I defined in step one."

Tips & gotchas

This skill is most effective when used after an iterative drafting process where instructions or feedback have changed. It requires the agent to maintain a clear record of the original constraints to perform an accurate comparison.