Research & Data

https://agentskills.io 準拠

Reusable Codex skills for automation, research, engineering, and practical workflows.

claude skill

Codex skill for searching, verifying, and summarizing academic papers into compact Zotero/arXiv-style tables.

package-vuln-audit is a portable Agent Skill for authorized software-package vulnerability discovery and analysis. It combines traditional static/dynamic tools, AI-assisted source reasoning, subagent delegation, validation evidence, CVSS scoring, and progressive disclosure.

Cursor agent skill for managing Azure Cosmos DB — accounts, databases, containers, documents, queries, throughput, indexing, stored procedures, triggers, UDFs, and more

开源意图搜索器：描述需求 → 自动拆解能力 → 跨 GitHub/npm/PyPI 多源搜索 → 评分评级 → 输出选型报告。可作为 Claude Skill 使用。

Public Codex skills for reusable bioinformatics analysis workflows

Personal Research Lab / 个人研究实验室 Research methods, deep work workflows, agent skills, and practical notes for building durable thinking systems.

Recently updated agent-skill-related GitHub repository: naoterumaker/japan-gyousei-data.

A small, AI-assisted pipeline that turns messy bank data into a clean finance database, then presents it as a dashboard. It is delivered as two Claude skills plus one optional app.

Codex skill for AI product PM brainstorming, architecture, prompt discipline, and GitHub prototype research

A reusable Codex skill for orchestrating source-backed industry and company research reports.

Reusable agent skill for building a personal LLM-study wiki (Karpathy LLM-wiki style): paper/book ingest discipline, bespoke-page templates, figure patterns, anti-patterns.

AI agent skill for weekly Go ecosystem radar reports — collects, curates, and generates self-contained HTML

A Codex skill for Feynman-style industry exploration with visual roadmaps, familiar examples, role-based analysis, and structured deliverables. 用费曼式讲解、探索路线图、熟悉案例和身份视角，帮助用户快速理解一个行业。

AI-powered project handover for Claude Code and Codex: onboarding docs plus privacy-filtered AI memories, verified against current code before import.

Bytesync Agent Skills Marketplace

Publication-grade meta-analysis & systematic review as an agent skill — run by Claude Code · Codex · any SKILL.md agent, with seven human sign-off gates and red lights computed by R. PRISMA 2020 · GRADE · AMSTAR 2.

Community agent skills, including model-radar for LLM model selection on OpenRouter.

From thought to skill. From signal to structure.

Turn "ask AI" into goal-driven learning — a roadmap that walks you through concepts and auto-builds your knowledge vault

Agent skills extracted from game-development work and related agentic-workflow research

A Codex Skill that turns company research into a multi-investor panel debate.