Browse the library

This library provides 177 validated skills and access to 100+ scientific databases to transform any AI agent into an AI scientist for research.

An autonomous MCP agent designed for deep web and local research, providing detailed, factual, and unbiased reports with citations for any given task.

This unofficial Python API and agentic skill provide programmatic access to Google Gemini Notebook features for AI agents like Claude Code.

A comprehensive open-source library of skills that empowers AI agents to autonomously conduct AI research, from ideation and experimentation to paper writing.

Enables scientific AI agents to perform reproducible research by running code, querying data, and generating reports from a local workbench.

This MCP server allows AI agents to generate in-depth research reports rapidly, supporting various LLMs and web search engines with local data processing.

Provides a vast collection of over 23,000 AI agent skills tailored for empirical research across various social science disciplines, aiding academic workflows.

hyperresearch

jordan-gibbs

This plugin transforms Claude Code or OpenAI Codex into a powerful deep research agent that collects, synthesizes, and stores web research into a persistent, searchable knowledge base.

This agent skill integrates the complete Traditional Chinese Medicine thinking system of Master Ni Haixia, allowing AI to perform diagnostics and herbal selections.

This MCP server allows AI agents to read arXiv papers, extract BibTeX, and keep topic watches, with papers stored locally.

This skill provides a structured, human-in-the-loop research workflow for AI agents to generate outlines and conduct deep investigations on various topics.

An Agent Skill for Claude Code/Codex that organizes and makes searchable Ni Haisha TCM course materials for study and reference.

A multi-stage skill for AI coding agents that automates the math modeling process, from analysis and coding in Python/MATLAB to generating DOCX competition papers.

Engram is an evidence-based learning plugin that turns AI agents into tutors to help humans learn and retain information.

This autonomous AI research engine conducts computational experiments, generates code, executes training, and produces full LaTeX papers with real experimental data.

Aggregates native AI search from platforms like Gemini and Grok, allowing agents to perform deep web searches and extract data from diverse sources.

This MCP server and Zotero plugin allow your AI assistants like Claude to deeply integrate with Zotero, enabling smart search, content extraction, and semantic analysis of your literature library.

An official Jina AI remote MCP server providing tools for AI agents to read web pages, search content, generate embeddings, and rerank results using Jina AI's APIs.

This plugin provides guardrails and skills for Claude Code to perform academic research tasks like paper reproduction, experiment design, and review, preventing common AI mistakes.

Audits autoresearch papers for self-consistency and fabrication using 61 integrity signals, providing a deterministic verdict.

This implements a Parallel-Agent Reinforcement Learning (PARL) paradigm for training AI models to decompose tasks into parallel subtasks and coordinate multiple agents.

This plugin integrates Google Antigravity with Claude Code for local document summarization, deep web research, and audio/video transcription.

A Claude Skill that organizes course materials, PDFs, and PPTs into Markdown/Obsidian revision notes.