skills.rest
raja21068
raja21068COMMUNITY
Explore all skills by this author →
WebsiteGitHub

auto-review-loop-minimax

Run iterative research paper review and fix loops with MiniMax.

2|1|Updated Apr 19, 2026
Search Tags:#state persistence#paper writing#research iteration#autonomous review#miniMax api#experiment verification#self-healing fixes
One-click install
npx skills add https://github.com/raja21068/AutoResearch --skill auto-review-loop-minimax-raja21068
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: auto-review-loop-minimax
Source: https://github.com/raja21068/AutoResearch/tree/main/skills/aris/auto-review-loop-minimax
Command: npx skills add https://github.com/raja21068/AutoResearch --skill auto-review-loop-minimax-raja21068

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It solves the problem of getting slow, inconsistent research revision cycles by repeatedly reviewing a draft, applying targeted fixes, and re-reviewing until an external evaluator is satisfied or a maximum number of rounds is reached.

Core Features & Use Cases

  • Autonomous multi-round review loop: Iteratively runs review → implements fixes → re-reviews up to MAX_ROUNDS (default 4).
  • MiniMax-based external reviewer (with fallback): Uses an MCP MiniMax chat tool when available, otherwise falls back to direct curl calls to the MiniMax API.
  • Persistent state and cumulative logging: Recovers from context compaction by saving and resuming loop state in review-stage/REVIEW_STATE.json and appending all round details to review-stage/AUTO_REVIEW.md.
  • Submission readiness gating: Stops early when the reviewer score and verdict indicate readiness (score >= 6/10 and verdict contains accept/sufficient/ready).

Quick Start

Trigger the skill with the instruction: "auto review loop minimax" to run an end-to-end review-and-fix cycle using MiniMax-M2.7 and persist results under review-stage/.

Frequently Asked Questions about auto-review-loop-minimax

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

FAQPage Schema
How do I automate paper revision and experiment verification with an external reviewer?▼

Automate paper revision by sending research context to MiniMax for external review, applying reviewer-directed fixes, and re-reviewing iteratively. This autonomous review loop runs up to four rounds, validating methods and strengthening results until submission readiness is achieved.

Can I use MiniMax API for iterative academic writeups and method validation?▼

Yes, you can use the MiniMax API for iterative academic writeups by running an autonomous multi-round review loop. It supports MCP or direct curl fallback to send research context, receive external feedback, and apply targeted fixes for method validation.

How does state persistence handle context compaction during research iteration?▼

State persistence handles context compaction by saving loop state to REVIEW_STATE.json and appending round details to AUTO_REVIEW.md. This allows the autonomous review loop to recover, resume, and continue applying self-healing fixes seamlessly after interruptions.

What stops the autonomous review loop from running indefinitely?▼

The autonomous review loop stops when the external reviewer score reaches 6/10 or higher with an accept, sufficient, or ready verdict. If submission readiness gating is not met, it halts after reaching the default maximum of four review rounds.

Do I need YAML frontmatter configuration to run an external review cycle?▼

Yes, YAML frontmatter configuration specifying a MiniMax model is required to run the external review cycle. This setup enables the system to connect to the API, execute reviewer-directed fixes, and persist logs for later resumption.

What is the best way to get consistent feedback for research revision cycles?▼

The best way to get consistent research revision feedback is using an autonomous multi-round review loop with a non-Codex external reviewer like MiniMax. It repeatedly reviews drafts, applies targeted fixes, and re-reviews until readiness criteria are met.

Related Skills

Explore More Education & Research Skills→
Mang30Mang30

auto-review-loop-minimax

Automate multi-round research review and iterative fixes via MiniMax API.

Community
Advanced
nahisahonahisaho
3

scientific-latex-export

Convert experimental results into LaTeX IMRaD manuscripts with BibTeX references.

Community
Intermediate
jkfeejkfee

auto-review-loop-minimax

Automate multi-round ML research review workflows using the MiniMax API.

Community
Advanced
dogekikidogekiki

auto-review-loop

Coordinate external reviewer backends with local implementation tasks in iterative research cycles.

Community
Advanced