# ARIS — Claude Code Research Skills (Markdown-Only)

> ARIS is a Markdown-only skill library for Claude Code: a research pipeline (lit search, idea discovery, audits) with optional Codex/Gemini add-ons.

## Install

Save the content below to `.claude/skills/` or append to your `CLAUDE.md`:

## Quick Use

```bash
git clone https://github.com/wanshuiyin/Auto-claude-code-research-in-sleep.git
mkdir -p ~/.claude/skills/
cp -r Auto-claude-code-research-in-sleep/skills/* ~/.claude/skills/
# Optional Codex MCP (from README):
npm install -g @openai/codex
claude mcp add codex -s user -- codex mcp-server
```

## Intro

ARIS is a Markdown-only skill library for Claude Code: a research pipeline (lit search, idea discovery, audits) with optional Codex/Gemini add-ons.

**Best for:** Teams who run repeated research workflows and want them standardized as skills

**Works with:** Claude Code skills, optional: Codex MCP, gemini-cli, Python extras

**Setup time:** 6-15 minutes

### Key facts (verified)

- GitHub: 9153 stars · 865 forks · pushed 2026-05-13.
- License: MIT · owner avatar + repo URL verified via GitHub API.
- README-verified entrypoint: `pip install vastai`.

## Main

- Treat it as a reproducible research pipeline: keep skills in version control and evolve them like code instead of re-prompting from scratch each week.

- Start with the core install (copy `skills/*` into `~/.claude/skills/`), then enable optional sources (Codex MCP, Gemini) only when you need deeper coverage.

- Use the project’s change log sections as a playbook to understand which skills are stable defaults vs. opt-in experimental routes.

### Source-backed notes

- README provides an install section that clones the repo and copies `skills/*` into `~/.claude/skills/` for Claude Code.
- README includes an optional “Set up Codex MCP” section with `npm install -g @openai/codex` and a `claude mcp add ...` command.
- README changelog-style entries document ongoing additions and routing updates for skills and sources.

### FAQ

- **Do I need Python to use ARIS?**: Many skills are Markdown-only; optional integrations may require Python packages as described in README.
- **Where do Claude Code skills live?**: README copies them into `~/.claude/skills/` (create it if missing).
- **How do I add Codex tools?**: Follow README’s Codex MCP setup: install `@openai/codex` and add it via `claude mcp add`.

## Source & Thanks

> Source: https://github.com/wanshuiyin/Auto-claude-code-research-in-sleep
> License: MIT
> GitHub stars: 9153 · forks: 865

---

<!-- ZH -->

## Quick Use

```bash
git clone https://github.com/wanshuiyin/Auto-claude-code-research-in-sleep.git
mkdir -p ~/.claude/skills/
cp -r Auto-claude-code-research-in-sleep/skills/* ~/.claude/skills/
# Optional Codex MCP (from README):
npm install -g @openai/codex
claude mcp add codex -s user -- codex mcp-server
```

## Intro

ARIS 是面向 Claude Code 的纯 Markdown 技能库：把文献检索、想法发散、审计与迭代等研究流程做成可复用技能；README 还给出可选的 Codex/Gemini 接入命令。

**Best for:** 需要反复跑研究流程、想把流程标准化成技能的团队

**Works with:** Claude Code skills；可选：Codex MCP、gemini-cli、Python 扩展依赖

**Setup time:** 6-15 minutes

### Key facts (verified)

- GitHub：9153 stars · 865 forks；最近更新 2026-05-13。
- 许可证：MIT；作者头像与仓库链接均已通过 GitHub API 复核。
- README 中核对过的入口命令：`pip install vastai`。

## Main

- 把它当作“可复现研究流水线”：技能进版本库，像代码一样迭代，而不是每次重写 prompt。

- 先完成核心安装（把 `skills/*` 拷到 `~/.claude/skills/`），需要更深覆盖再启用 Codex/Gemini 等可选来源。

- 用 README 的变更记录当作路线图：区分稳定默认与实验性技能，避免一上来全开导致成本与不确定性飙升。

### Source-backed notes

- README 的安装章节：git clone 后将 `skills/*` 拷贝到 `~/.claude/skills/` 给 Claude Code 使用。
- README 提供可选的 Codex MCP 接入：`npm install -g @openai/codex` 并用 `claude mcp add ...` 添加服务器。
- README 以变更记录方式持续更新技能与路由策略，方便追踪默认行为与新能力。

### FAQ

- **必须装 Python 吗？**：不少技能是纯 Markdown；但部分可选集成按 README 可能需要 Python 包。
- **Claude Code 的 skills 放哪？**：按 README：复制到 `~/.claude/skills/`（不存在就先创建）。
- **怎么接入 Codex 工具？**：按 README：安装 `@openai/codex`，再用 `claude mcp add` 添加 Codex MCP。

## Source & Thanks

> Source: https://github.com/wanshuiyin/Auto-claude-code-research-in-sleep
> License: MIT
> GitHub stars: 9153 · forks: 865


---
Source: https://tokrepo.com/en/workflows/aris-claude-code-research-skills-markdown-only
Author: Skill Factory