skills.rest
KwongFuk
KwongFukCOMMUNITY
Explore all skills by this author →
GitHub

auto-review-loop-llm

Automate iterative research review loops with OpenAI-compatible LLMs.

Updated Apr 10, 2026
Search Tags:#research#llm#mcp#openai#review-loop#autonomous-review
One-click install
npx skills add https://github.com/KwongFuk/codex-skills --skill auto-review-loop-llm-kwongfuk
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: auto-review-loop-llm
Source: https://github.com/KwongFuk/codex-skills/tree/main/global/auto-review-loop-llm
Command: npx skills add https://github.com/KwongFuk/codex-skills --skill auto-review-loop-llm-kwongfuk

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Autonomously iterate: review → implement fixes → re-review, until the external reviewer gives a positive assessment or MAX_ROUNDS is reached.

Core Features & Use Cases

  • Autonomous review loop with configurable LLM provider
  • State persistence via REVIEW_STATE.json and auto log to AUTO_REVIEW.md
  • MCP server integration for chat-based reviews

Quick Start

Trigger the autonomous review loop by issuing the command 'auto review loop llm' or 'llm review' while using an OpenAI-compatible LLM via the llm-chat MCP server.

Frequently Asked Questions about auto-review-loop-llm

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

FAQPage Schema
How do I automate multi-round LLM review for research work?▼

Automate multi-round LLM review by triggering an autonomous loop that iteratively critiques research, suggests fixes, and re-evaluates until a positive verdict or max rounds. It logs progress to AUTO_REVIEW.md.

Does the autonomous review loop work with an OpenAI-compatible LLM?▼

Yes, the autonomous review loop works with any OpenAI-compatible LLM. It supports direct API calls and MCP-based chat integration, allowing flexible LLM provider configuration for iterative critiques.

How does state persistence work during iterative LLM critique?▼

State persistence works by saving review loop progress to REVIEW_STATE.json. This ensures iterative critique cycles are tracked and resumable across sessions until the review concludes.

Can I use MCP server integration for chat-based research reviews?▼

Yes, you can use MCP server integration for chat-based research reviews. Trigger the loop via commands like 'auto review loop llm' while using the llm-chat MCP server for structured, multi-round critiques.

What are the limitations of autonomous review loops for research projects?▼

The main limitation of autonomous review loops is the MAX_ROUNDS threshold, which caps iterative critiques and fixes. The loop terminates if a positive verdict is not reached within this maximum.

Related Skills

Explore More Education & Research Skills→
jandan138jandan138

auto-review-loop-llm

Automates an end-to-end ML research review loop with cross-model critique and iterative fixes.

Community
Advanced
shishishishi
3

codex-review

Automate iterative code reviews with Codex and Claude Code patches.

Community
Advanced
nickpeterson92nickpeterson92
2

websocket-protocol

Documents WebSocket message formats for chat streaming between clients and servers.

Community
Basic
KYRIE66nbKYRIE66nb

auto-review-loop-llm

Automate iterative ML research reviews using OpenAI-compatible LLM APIs.

Community
Advanced