auto-review-loop-minimax

Automating academic review involves orchestrating a multi-round loop where an API reviews content and implements fixes until passing. This Skill manages that loop automatically, handling state persistence and documenting each round's assessments and corrective actions.

Yes, you can use the MiniMax API for external review through either MCP Tool or curl methods. The automation loop submits your research content to the API, receives feedback, implements fixes, and re-reviews until a positive assessment is achieved or rounds max out.

To start an automated research review loop, trigger the system using the phrase "auto review loop minimax" or "minimax review" followed by the specific topic or scope you wish to review.

The automated review loop recovers from context compaction using built-in state persistence. This mechanism saves the progress of your research review rounds, allowing the iterative implementation and feedback cycle to resume seamlessly without losing previous assessments.

The maximum number of review rounds is configured to prevent infinite loops during iterative research fixes. The automation continues the cycle of review and implementation until a positive assessment is received or this maximum round limit is reached.

You need curl and json dependencies to run the MiniMax API review automation if using the curl method. The Skill also supports the mcp__minimax-chat__minimax_chat component, offering an alternative integration path for the external review process.