Agent skill · research science · aradotso
aris-autonomous-research
aris-autonomous-research — an agent skill for coding agents.
Why this skill is useful
Adds executable workflows for autonomous ML research, including idea discovery, experiment automation, and paper writing that aren't available in public documentation.
What it needs
Requires git, python installed locally. Requires anthropic, openai, wandb 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 Option A: copy skills to your project cp -r skills/ /your/project/.claude/skills/ Option B: symlink (keeps skills up to date) ln -s /path/to/Auto-claude-code-research-in-sleep/skills /your/project/.claude/skills bash cd mcp-servers/llm-chat pip install -r requirements.txt # or: uv pip install -r requirements.txt json { "mcpServers": { "llm-chat": { "command": "python", "args": ["/path/to/Auto-claude-code-research-in-sleep/mcp-servers/llm-chat/server.py"], "env": { "OPENAIAPIKEY": "$OPENAIAPIKEY", "LLMMODEL": "gpt-4o" } } } } json { "mcpServers": { "codex": { "command": "npx", "args": ["@openai/codex-mcp"], "env": { "OPENAIAPIKEY": "$OPENAIAPIKEY" } } } } /research-pipeline "factorized gap in discrete diffusion LMs" /research-pipeline "improve positional encoding in transformers" — ref paper: https://arxiv.org/abs/2104.09864, base repo: https://github.com/facebookresearch/fairseq /idea-discovery "sparse attention in long-context LLMs" /experiment-bridge "ideaproposal.md" — base repo: https://github.com/huggingface/diffusers python experimentresults/run001/train.py (auto-generated by /experiment-bridge) import wandb import torch from torch.utils.data import DataLoader wandb.init( project=os.environ.get("WANDBPROJECT", "aris-experiment"), config={ "method": "factorizeddiscretediffusion", "lr": 3e-4, "epochs": 50, "batchsize": 32, } ) for epoch in range(config.epochs): for batch in dataloader: loss = model(batch) wandb.log({"loss": loss.item(), "epoch": epoch}) /literature-review "discrete diffusion language models" /paper-write "experimentresults/ + ideaproposal.md" — venue: NeurIPS /rebuttal "paper/ + reviews/" — venue: ICML, character limit: 5000 /paper-slides "paper/" # Beamer PDF + PPTX + speaker notes + Q&A prep /paper-poster "paper/" # A0/A1 poster PDF + PPTX + SVG (venue colors) bash Kimi as reviewer export LLMBASEURL="https://api.moonsh …
How to use it
Reference it in AdaL, Claude Code, Cursor or any coding agent — nothing to install:
@skills aradotso/aris-autonomous-research