ARS-Codex

ARS-Codex is a separate Codex release by Cheng-I Wu (GitHub user Imbad0202), built from the same author's Academic Research Skills, which is also listed on this site. It packages the original suite as one Skill called academic-research-suite, which routes your request to research, writing, review, pipeline or systematic-review screening based on what you ask. Its version number follows the original suite, and the README records the exact source repository and commit. What sets it apart is that it was reworked to run the Codex way: by default it uses Codex's own sub-agents, and the fixed team mode and hooks only take effect if you turn them on. The README also says plainly that screening accuracy and model effectiveness under Codex haven't been measured yet.

Who it's for:researchers who mainly use Codex for research and paper writing. If you use Claude Code, see the original edition.

Before you start: install it with Codex's plugin marketplace commands and restart your session. Normal use doesn't need the full runtime mode or hooks, and we suggest leaving them off, which is the default. Citation checks send reference details to scholarly indexes; cross-checking with other models needs your explicit consent. The suite includes material for an experiment agent, which doesn't mean it runs experiments by default. The license is CC BY-NC 4.0, for non-commercial use only. Codex subscription or model costs are separate.

Based on the project's public documentation. Not tested by AI Inspiration World.