wanshuiyinCommunity·89 Agent Skills Included

Auto-claude-code-research-in-sleep

Automate the full academic research pipeline from idea to submission

Automates the entire research lifecycle: literature search, idea generation, novelty checks, experiment planning, GPU job orchestration, paper writing, proof verification, and citation audits. Eliminates manual literature reviews, hallucinated citations, unverified claims, and wasted GPU hours on broken training runs. Uses cross-model review loops so one AI writes and another independently audits, catching errors before submission.

15,551|1,361

↗

npx skills add wanshuiyin/Auto-claude-code-research-in-sleep --all -g -y

All Skills in This Repository (89)

Pure Emerald Level Indicators

📦 In Repo

wanshuiyin

paper-write

Generate LaTeX paper sections from an outline and narrative report.

proof-writer

Generate rigorous mathematical proofs for ML/AI theory claims.

research-refine-pipeline

Orchestrate research workflows from method refinement to experiment planning.

auto-review-loop-llm

Automate iterative research review loops using an OpenAI-compatible LLM API.

experiment-plan

Generates a claim-driven experiment roadmap from a research proposal or method idea.

idea-creator

Generate and rank research ideas with novelty and feasibility checks.

monitor-experiment

Monitor remote machine learning experiments via SSH and collect JSON results.

research-pipeline

Automate machine learning research from idea generation to submission-ready papers.

research-refine

Refine vague research directions into problem-anchored method plans.

idea-discovery

Automates research idea discovery from literature surveys to validated experiment plans for ML workflows.

paper-compile

Compile LaTeX papers to PDF with latexmk and fix common errors.

analyze-results

Analyze ML experiment results from JSON/CSV to compute statistics and generate insights.

Community

Intermediate

Frequently Asked Questions

FAQPage Schema

How to install ARIS?▼

Run `npx skills add wanshuiyin/Auto-claude-code-research-in-sleep --all -g -y` in your terminal to install all skills in this suite globally.

What does ARIS actually do?▼

It chains research skills into full pipelines: find an idea, verify novelty, plan and run experiments, write the paper, and audit every claim and citation before submission.

Which AI agents does ARIS work with?▼

It runs in Claude Code, Codex CLI, Cursor, Trae, Antigravity, GitHub Copilot CLI, and OpenClaw, using the standard SKILL.md format.

How does ARIS prevent fabricated results?▼

A separate model family audits every number, citation, and proof against raw result files, so the writing agent never grades its own work.

Can ARIS run experiments on cloud GPUs?▼

Yes. Built-in skills rent and manage GPUs on vast.ai, Modal, and SSH servers, with cost estimates, OOM retry, and automatic cleanup.