skills.rest
OpenLAIR
OpenLAIROFFICIAL
Explore all skills by this author →
GitHub

aris-infra

Configures MCP servers and Python tools for cross-model adversarial research review.

1.1k|116|Updated Feb 26, 2026
Search Tags:#research automation#mcp servers#LLM configuration#cross-model review#Arxiv#Setup#Claude code
One-click install
npx skills add https://github.com/OpenLAIR/dr-claw --skill aris-infra
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: aris-infra
Source: https://github.com/OpenLAIR/dr-claw/tree/main/skills/aris-infra
Command: npx skills add https://github.com/OpenLAIR/dr-claw --skill aris-infra

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires httpx, arxiv, requests, @openai/codex, and includes scripts (resource) and references (resource) components.

What problem does it solve?

Setting up an AI research assistant that uses cross-model adversarial review requires manually registering multiple MCP servers, installing Python dependencies, and validating API keys, which is error-prone and tedious. This Skill automates the entire ARIS infrastructure setup so Claude Code can execute research tasks while an external LLM (GPT, Gemini, MiniMax) provides independent critical review.

Core Features & Use Cases

  • One-Command Setup: An interactive setup.sh script checks prerequisites, installs Python dependencies (httpx, arxiv, requests), symlinks ARIS skills into ~/.claude/skills, and registers your chosen MCP reviewer server.
  • Multiple Reviewer Backends: Register Codex (GPT), a generic OpenAI-compatible LLM bridge, Gemini review, Claude cross-session review, MiniMax chat, or Feishu/Lark notifications as MCP servers.
  • Bundled Research Tools: Includes arXiv search/download, Semantic Scholar fetching with citation filters, a persistent research wiki, and a GPU training/download watchdog daemon.
  • Use Case: A researcher wants Claude Code to run experiments overnight while GPT-5 reviews the results. They run the setup script, select the Codex backend, set their OPENAI_API_KEY, and immediately gain access to the full ARIS workflow skills (idea discovery, experiment bridge, paper writing).

Quick Start

Run bash skills/aris-infra/setup.sh and follow the interactive prompts to install dependencies and register a reviewer MCP server.

Frequently Asked Questions about aris-infra

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

FAQPage Schema
How do I set up cross-model LLM review in Claude Code?▼

Run the ARIS setup script with bash skills/aris-infra/setup.sh and choose a reviewer backend such as Codex, a generic OpenAI-compatible API, or Gemini. The script registers the MCP server with claude mcp add and verifies your API key environment variables.

What MCP servers can I register for adversarial research review?▼

ARIS provides five MCP servers: codex for GPT-based review, llm-chat for any OpenAI-compatible API, gemini-review for Google Gemini, claude-review for cross-session Claude review, and minimax-chat for MiniMax. An optional feishu-bridge server sends Lark notifications.

Does ARIS work with Gemini instead of OpenAI models?▼

Yes, register the gemini-review MCP server and set GEMINI_API_KEY or GOOGLE_API_KEY in your environment. The default model is gemini-2.5-pro, configurable via GEMINI_REVIEW_MODEL, and it supports both API and Gemini CLI backends.

Why is my MCP server not found after setup?▼

The MCP server registration must use the -s user flag with claude mcp add, and you need to restart your Claude Code session for the server to take effect. Verify registration with claude mcp list.

What Python dependencies does ARIS require?▼

ARIS requires Python 3.10 or later plus the httpx, arxiv, and requests packages, installed via pip install httpx arxiv requests. These power the bundled arXiv fetcher, Semantic Scholar client, research wiki, and watchdog daemon.

Related Skills

Explore More Education & Research Skills→
rohanbhatia27rohanbhatia27
1

second-opinion

Run external LLM code reviews on uncommitted changes, branch diffs, or commits.

Community
Intermediate
wanshuiyinwanshuiyin
15.6k

research-pipeline

Orchestrates end-to-end autonomous research from idea discovery through experiments to paper PDF.

Community
Advanced
Mang30Mang30

research-wiki

Ingests papers and records typed relationships in a graph-based research wiki.

Community
Advanced
kwt00kwt00

mcp-setup

Configure MCP servers for Grok Build via the `claude mcp add` CLI.

Community
Intermediate