wanshuiyin — 7 Open-Source AI Agent Skills

wanshuiyin is the author of 7 open-source AI agent skills and MCP servers spanning mcp-server, codex-skill, claude-skill, agent-tool, with a combined 20,212+ GitHub stars. On AgentSkillsHub each is quality-scored (avg 60/100) and security-graded — 7 verified safe.

Top skills by wanshuiyin

• Auto-claude-code-research-in-sleep — ARIS ⚔️ (Auto-Research-In-Sleep) — Lightweight Markdown-only skills for autonomous ML research: cross-model review loops, idea discovery, an (18.5k★ · quality 64/100 · security: safe)
• HERO-Anti-OverDefense — HERO = Hashing · Edge cases · Rubrics · Overbuild — the four shapes coding agents over-defend in. A paste-in contract that stops them. Works (741★ · quality 60/100 · security: safe)
• ARIS-in-AI-Offer — Bilingual (中文+EN) ML / LLM / diffusion / agent interview cheat sheets for AI 秋招 — generated by ARIS /interview-cheatsheet, rendered by /rend (627★ · quality 58/100 · security: safe)
• Anti-Autoresearch — Don't trust an autoresearch paper at face value. Reviewer-side integrity forensics (self-consistency + fabrication), deterministic verdict. (172★ · quality 60/100 · security: safe)
• ARIS-Movie-Director — Agentic, long-horizon visual generation: a fuzzy story → a cross-model-audited image-based movie. Brings ARIS's research-wiki + multi-agent (74★ · quality 60/100 · security: safe)
• ARIS-Anything — Generalizing ARIS from 科研 to 研究 — same five-step loop (plan · draft · adversarial review · iterate · persist) applied beyond academic resear (55★ · quality 56/100 · security: safe)
• ALIGN-Agentic-Loop-Image-GeneratioN — Agentic loop image generation without diffusion or autoregressive image models. Coding agents draw in p5.js, using iterative refinement and (45★ · quality 65/100 · security: safe)

View wanshuiyin on GitHub · Explore audited agent skills