
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。
Sign in to rate this tool.
Overview

Agentero is a desktop‑first research workspace that lets you keep papers, notes and the surrounding context in a local vault while interacting with an AI agent of your choice. It positions itself as a bridge between traditional reference managers and AI‑augmented reading, offering import from Zotero and export to BibTeX formats.
The tool is aimed at researchers who already maintain a bibliography in Zotero or who use BibTeX‑based workflows. You can migrate an existing Zotero library with tags, notes and attachments intact, then continue to add new papers via RSS feeds, arXiv searches or a browser connector that drops a PDF into the vault with a single click. Once a PDF is in the vault, Agentero automatically extracts figures, tables and equations, turning them into searchable structures that appear in a side pane. Selecting any fragment, such as a formula or paragraph, opens that side panel where a locally connected agent, linked through the ACP protocol, answers questions using the full paper context and can cite specific sections.
Agentero stores all files, notes and the evolving knowledge graph on your machine; no data leaves the vault unless you explicitly invoke a third‑party service such as DOI lookup, a translation API, or an agent that you have logged in locally. The knowledge graph grows as you create double‑linked markdown notes; each note can contain [[wikilink]] references that automatically surface related papers, and the graph can be exported as BibTeX or BibLaTeX for use in other citation tools. A command‑line interface lets you script vault operations, adding, tagging or exporting papers, while remote access is possible through an SSH tunnel that presents the vault as if it were local, enabling the same workflow on a separate machine.
The application runs on macOS, Windows and Linux, with additional packages for Android and an iOS TestFlight build. Installers are available as native binaries, AppImage.deb and .rpm packages, and the software can also be installed via Homebrew. The project is open source under the MIT license, and the source code and release assets are hosted on GitHub.
Key features
- One‑click import from Zotero preserving tags, notes and attachments
- Automatic detection of figures, tables and equations inside PDFs, displayed as searchable objects
Details
- Pricing
- Free
- Category
- Code
Suggest an edit
Sign in to suggest an edit.
Pros and cons
Pros
- Preserves existing Zotero workflow while adding AI‑driven reading
- Local‑first design ensures research data stays on your device unless you choose otherwise
- Supports a range of agents without binding to a specific model or cloud provider
Cons
- No pricing information is published; the listing does not indicate whether any paid tiers exist
- Requires you to run or install a compatible local agent for AI features
- Advanced features such as translation or DOI lookup depend on external services you must configure yourself

