Agent skill · research science · aradotso
aris-autonomous-ml-research
aris-autonomous-ml-research — an agent skill for coding agents.
Why this skill is useful
Adds executable workflows for autonomous ML research, including experiment automation and paper writing that extend the AI's capabilities.
What it needs
Requires @anthropic-ai/claude-code, @openai/codex installed locally. Requires anthropic, openai account access. About 7k tokens when loaded. Last updated 2026-07-09. 70 stars on the source repository.
What this skill does
bash git clone https://github.com/wanshuiyin/Auto-claude-code-research-in-sleep.git cd Auto-claude-code-research-in-sleep bash npm install -g @anthropic-ai/claude-code bash npm install -g @openai/codex json { "mcpServers": { "codex": { "command": "codex", "args": ["mcp"], "env": { "OPENAIAPIKEY": "$OPENAIAPIKEY" } } } } bash Copy all skills to Claude Code's custom skills directory cp -r skills/claude-code/ ~/.claude/skills/ Or symlink to stay up to date ln -s $(pwd)/skills/claude-code ~/.claude/skills/aris bash Required for Claude Code export ANTHROPICAPIKEY=youranthropickey Required for cross-model review (GPT-5.4 as reviewer) export OPENAIAPIKEY=youropenaikey Optional: alternative reviewer models (no OpenAI needed) export LLMREVIEWERBASEURL=https://api.minimax.chat/v1 export LLMREVIEWERAPIKEY=yourminimaxkey export LLMREVIEWERMODEL=MiniMax-M2.7 json { "mcpServers": { "llm-chat": { "command": "node", "args": ["mcp-servers/llm-chat/index.js"], "env": { "LLMBASEURL": "$LLMREVIEWERBASEURL", "LLMAPIKEY": "$LLMREVIEWERAPIKEY", "LLMMODEL": "$LLMREVIEWERMODEL" } } } } /research-pipeline "factorized gap in discrete diffusion LMs" /research-pipeline "improve method X" — ref paper: https://arxiv.org/abs/2406.04329, base repo: https://github.com/org/project /research-pipeline "topic" — ref paper: <arxivurl> # Optional: paper to improve — base repo: <githuburl> # Optional: codebase to build on — venue: ICML # Target venue (default: ICML) — compact: true # Lean summaries for short-context models /idea-discovery "discrete diffusion language models" /experiment-bridge "run ablation on temperature scaling" — code review: true python Example: what experiment-bridge automates 1. Claude Code writes training script 2. GPT-5.4 reviews the code (code review gate) 3. If approved → submits to GPU cluster 4. …
How to use it
Reference it in AdaL, Claude Code, Cursor or any coding agent — nothing to install:
@skills aradotso/aris-autonomous-ml-research