Versions:
Agentero, published by poco-ai, is a literature management tool designed to be agent-friendly and agent-native, addressing limitations of traditional reference managers that were built primarily for human users. The software addresses three core problems: reading highlights and notes are typically locked inside individual files, making it difficult for agents to reuse knowledge across papers; each new conversation requires context to be re-supplied because there is no stable local knowledge map; and PDFs, while convenient for human reading, are not an ideal format for agents to consume. Agentero explores collaborative workflows between people and agents in academic literature management. It connects to local agents through ACP (Bring Your Own Agent), and supports word-selection-based dialogue, paper import, and skill import. The tool is compatible with the Zotero ecosystem and provides features including paper translation, bidirectional linking with knowledge graphs, and deep PDF reading. These capabilities position it within the reference management and research productivity category, aimed at researchers and knowledge workers who want to integrate AI agents directly into their literature workflows. The current version is 0.10.2, and the project has released 14 versions to date, indicating active ongoing development. By building a persistent, locally stored knowledge map, Agentero allows agents to maintain context across sessions and reuse annotations and notes from multiple papers, rather than starting from scratch in every interaction.
Tags: