wanshuiyin / Auto-claude-code-research-in-sleep

Auto-claude-code-research-in-sleep (ARIS ⚔️) 中文版 README | English > 🌙 **Let Claude Code do research while you sleep.** Wake up to find your paper scored, weaknesses identified, experiments run, and narrative rewritten — autonomously. · · -orange?style=flat) · 💬 Join Community · Custom Claude Code skills for autonomous ML research workflows. These skills orchestrate **cross-model collaboration** — Claude Code drives the research while an external LLM (via Codex MCP) acts as a critical reviewer. 🔀 **Also supports alternative model combinations (GLM, MiniMax, Kimi, LongCat, DeepSeek, etc.) — no Claude or OpenAI API required.** > 💭 **Why not self-play with a single model?** Using Claude Code subagents or agent teams for both execution and review is technically possible, but tends to fall into **local minima** — the same model reviewing its own patterns creates blind spots. > > *Think of it like adversarial vs. stochastic bandits: a single model self-reviewing is the stochastic case (predictable reward noise), while cross-model review is adversarial (the reviewer actively probes weaknesses the executor didn't anticipate) — and adversarial bandits are fundamentally harder to game.* > > 💭 **Why two models, not more?** Two is the minimum needed to break self-play blind spots, and 2-player games converge to Nash equilibrium far more efficiently than n-player ones. Adding more reviewers increases API cost and coordination overhead with diminishing returns — the biggest gain is going from 1→2, not 2→4. > > Claude Code's strength is fast, fluid execution; Codex (GPT-5.4 xhigh) is slower but more deliberate and rigorous in critique. These complementary styles — **speed × rigor** — produce better outcomes than either model talking to itself. 📢 What's New • **2026-03-15** — 🔀 **Bring your own model!** Any OpenAI-compatible API now works as reviewer via MCP server. GLM, MiniMax, Kimi, LongCat, DeepSeek all tested — **zero Claude or OpenAI API needed** • **2026-03-15** — 🐾 **OpenClaw adaptation guide** — use ARIS research workflows in OpenClaw without Claude Code slash skills • **2026-03-15** — 📐 ** ** — community skill for rigorous theorem proof drafting. 📚 **Anti-hallucination citations** — now fetches real BibTeX from DBLP/CrossRef instead of LLM-generated entries — on by default, zero install • **2026-03-14** — 📱 Feishu/Lark integration: three modes (off/push/interactive), mobile notifications for experiments, reviews, and checkpoints • **2026-03-13** — 🛑 Human-in-the-loop: configurable checkpoints across all workflows. Full autopilot or step-by-step approval • **2026-03-12** — 🔗 Zotero + Obsidian + local PDFs + arXiv/Scholar: multi-source literature search with cross-model novelty verification • **2026-03-11** — 🚀 Three end-to-end workflows complete: one prompt → top-venue-style paper. chains idea discovery → auto review → paper writing autonomously • **2026-03-09** — 📝 workflow: narrative report → structured outline → figures → LaTeX → compiled PDF → 2-round auto-improvement (4/10 → 8.5/10) 🚀 Quick Start > **Tip:** All pipeline behaviors are configurable via inline overrides — append to any command: > > | Parameter | Default | What it does | > |-----------|---------|-------------| > | | | Auto-continue at idea selection gate. Set to manually pick which idea to pursue before committing GPU time | > | | | Pause after each review round so you can read the score, give custom modification instructions, skip specific fixes, or stop early | > | | | Download top relevant arXiv PDFs during literature survey. When , only fetches metadata (title, abstract, authors) | > | | | Fetch real BibTeX from DBLP/CrossRef instead of LLM-generated entries. Eliminates hallucinated citations. Zero install | > > > **Important:** Codex MCP uses the model from , not from skill files. Make sure it says (recommended). Other options: , , . Run or edit the file directly. See full setup guide for details and alternative model combinations if you don't have Claude/OpenAI API. ✨ Features • 📊 **20 composable skills** — mix and match, or chain into full pipelines ( , , , ) • 🔍 **Literature & novelty** — multi-source paper search (**Zotero** + **Obsidian** + **local PDFs** + arXiv/Scholar) + cross-model novelty verification • 💡 **Idea discovery** — literature survey → brainstorm 8-12 ideas → novelty check → GPU pilot experiments → ranked report • 🔄 **Auto review loop** — 4-round autonomous review, 5/10 → 7.5/10 overnight with 20+ GPU experiments • 📝 **Paper writing** — narrative → outline → figures → LaTeX → PDF → auto-review (4/10 → 8.5/10), one command • 🤖 **Cross-model collaboration** — Claude Code executes, GPT-5.4 xhigh reviews. Adversarial, not self-play • 📝 **Peer review** — review others' papers as a conference reviewer, with structured scoring and meta-review • 🖥️ **GPU deployment** — auto rsync, screen sessions, multi-GPU parallel experiments, live monitoring • 🔀 **Flexible models** — default Claude × GPT-5.4, also supports GLM, MiniMax, Kimi, LongCat, DeepSeek, etc. — no Claude or OpenAI API required • 🛑 **Human-in-the-loop** — configurable checkpoints at key decisions. for full autopilot, to approve each step • 📱 **Feishu/Lark notifications** — three modes: **off (default, strongly recommended for most users)**, push-only (webhook, mobile alerts), interactive (approve/reject from Feishu). Zero impact when unconfigured Preview: Push cards (group) & Interactive chat (private) **Push Only** — group chat cards (experiment done, checkpoint, error, pipeline complete): **Interactive** — private chat with Claude Code (approve/reject, custom instructions): • 🧩 **Extensible** — domain-specific skills welcome! Add a and open a PR. See community skills like (architecture/EDA) --- 📈 Score Progression (Real Run) A real overnight 4-round run on an ML research project, from borderline reject to submission-ready: | Round | Score | What Happened | |-------|-------|---------------| | Init…