Agentero

Unified research workspace for humans & agents — PDF Q&A with citations, Vault, writing. 为人与 Agent 协作的科研工作台;补齐 Zotero 时代的人机上下文断点。

What is Agentero?

Unified research workspace for humans & agents — PDF Q&A with citations, Vault, writing. 为人与 Agent 协作的科研工作台;补齐 Zotero 时代的人机上下文断点。

Overview of Agentero

What it is

Agentero is an open-source unified research workspace designed for collaboration between humans and AI agents. It provides a structured environment for reading PDFs, taking notes, and managing literature, with built-in support for agent protocols. The project is licensed under MIT and is available on GitHub.

Key features

  • Structured PDF reading with outline navigation, highlighting, annotations, and full-text search. Automatic extraction of figures, tables, and mathematical formulas into Markdown and LaTeX.
  • Bidirectional linked notes using a WYSIWYG Markdown editor with [[wiki]] syntax, LaTeX rendering, and slash commands. Notes are natively linked to literature items.
  • Multi-mode translation service supporting word-level, page-by-page, and full-document translation. Built-in translation channels and optional custom API key (BYOK).
  • Open Agent protocol (ACP) to connect local agents (e.g., Claude Code, OpenCode, Codex). Also serves as an MCP server for external AI clients like ChatGPT Web and Claude Desktop.
  • Academic radar with multi-channel literature discovery: integration with Cool Papers, ModelScope community, RSS subscriptions, and personalized arXiv recommendations based on local library content.
  • Zotero database import (tags, notes, attachments), BibTeX/BibLaTeX export, and transparent local folder storage (PDF and Markdown files). Supports S3-compatible cloud sync and SSH tunnel remote access.

Use cases

  • Researchers who want to read PDFs with AI assistance while keeping all notes and context in one local workspace.
  • Users who need to batch-import literature from Zotero or web sources and automate metadata extraction and note-taking via CLI scripts.
  • Teams or individuals who want to connect local AI agents to a structured literature database for question answering, summarization, and citation analysis.
  • Anyone who needs a self-hosted, privacy-preserving research environment that works offline and syncs across devices via S3.

Getting started

Desktop clients are available for macOS, Windows, and Linux from the official website or GitHub Releases. On macOS, install via Homebrew: bash brew tap poco-ai/agentero brew install --cask agentero For the command-line tool (CLI): bash brew tap poco-ai/agentero brew install agentero agentero --help For detailed Linux dependencies (Ubuntu 22.04+ with webkit2gtk 4.1), see the repository README.

License and pricing

The project is open source under the MIT License. Pricing is not stated; the repository is marked as open source.

Based on the project's public README and repository metadata.

Add this badge to your README

Featured on YunoDB Badge styles & help

Maintain this project? Show that it is featured on YunoDB.

Markdown[![Featured on YunoDB](https://yunodb.com/badge/poco-aiagentero.svg)](https://yunodb.com/tools/poco-aiagentero?utm_source=github&utm_medium=badge)
HTML<a href="https://yunodb.com/tools/poco-aiagentero?utm_source=github&utm_medium=badge"><img src="https://yunodb.com/badge/poco-aiagentero.svg" alt="Featured on YunoDB" height="20"></a>