ATA – open-source terminal research agent for keeping up with papers

Name: ATA – open-source terminal research agent for keeping up with papers
Availability: InStock
Author: nimanima11

by nimanima11·Feb 25, 2026·3 points·1 comment

Visit Project View on HN

AI Analysis

●●SolidNiche GemSolve My Problem

Research agent with citation graphs and Zotero sync, but the core is orchestrating existing APIs.

Strengths

•Citation graph traversal + Zotero integration actually solves a real researcher pain (keeping up with papers).
•Multi-source search (arXiv, Semantic Scholar, OpenAlex, patents) beats single-platform alternatives.
•Persistent knowledge base that accumulates across sessions means it learns what you care about.

Weaknesses

•Built on top of OpenAI Codex CLI—mostly orchestrating existing tools, not novel architecture.
•Free/open but depends on external API calls (Semantic Scholar, arXiv). No evidence it's faster/better than browsing directly.

Post Description

Research is moving faster than ever across every field. We built ATA to help keep up.

ATA is a free, open-source fork of OpenAI's Codex CLI that adds a research stack: paper search (Semantic Scholar, arXiv, OpenAlex), citation graph traversal, PDF reading and synthesis, Hacker News discussion analysis, patent search (90+ offices via EPO), Zotero integration, and a persistent knowledge base that accumulates across sessions.

We also built a reading view for long-form output — when ATA synthesizes a paper or generates a report, it opens as a navigable document with sections and foldable details instead of dumping it into chat. Follow-up questions rewrite the relevant section in place rather than appending at the bottom.

Free, open source, runs locally.

Try it: npm install -g @a2a-ai/ata

https://github.com/Agents2AgentsAI/ata/