Imbad0202/academic-research-skills-codex

Codex-native Academic Research Skills suite for human-in-the-loop academic research workflows

Overview

ARS‑Codex is a Codex‑compatible skill that bundles the Academic Research Skills (ARS) workflow suite originally built for the Claude Code model. It repackages the same set of research‑oriented agents, prompts, and pipelines so they can be invoked from a Codex session via a single entry point – $academic‑research‑suite.

The package does not contain a new AI model; it merely adapts existing prompts, scripts, and runtime policies to work with the Codex plugin system. Once installed, users can ask Codex to plan systematic literature reviews, draft papers, run a full research‑to‑paper pipeline, screen study records, or design reproducibility experiments.

What the repository contains

Directory	Purpose
`skills/academic-research-suite/`	The core ARS skill – manifest, SKILL.md, and the `agents/`, `hooks/`, `scripts/` that drive the workflows.
`ars/` sub‑folders (`deep‑research/`, `academic‑paper/`, …)	Prompt‑templates and command‑recipes for each of the six main workflows (deep‑research, paper drafting, reviewer, pipeline, screener, experiment).
`codex/`	Codex‑specific adapter files: runtime manifest, optional full‑runtime topology, and policy documents (`model‑runtime‑policy.md`, audits).
`CODEX_FULL_RUNTIME_ADAPTER.md`	Documentation for the optional “full‑runtime” mode that enables a fixed agent‑team topology and hook pack.
`CHANGELOG.md`, `VERSION`	Version tracking – current release is v3.23.0 (mirrors the upstream ARS suite).
Installation scripts	Helper scripts for installing the skill directly via the Codex CLI or through the Codex Marketplace.

How to get it running

Via the Codex Marketplace (recommended):
```
codex plugin marketplace add Imbad0202/academic-research-skills-codex --ref main
codex plugin add ars-codex@ars-codex
```
This registers the plugin and makes the $academic‑research‑suite skill available.

Direct skill install (if you prefer a manual approach):

python3 "$HOME/.codex/skills/.system/skill-installer/scripts/install-skill-from-github.py" \
  --repo Imbad0202/academic-research-skills-codex \
  --ref main \
  --path skills/academic-research-suite \
  --method git

After installation, start a fresh Codex conversation and run /skills – you should see a single entry named academic‑research‑suite (or ARS‑Codex).

Updating – run the marketplace upgrade command or delete the skill folder and reinstall.

Main capabilities (the six workflows)

Workflow	When to use it	Example prompt you can give to Codex
deep‑research	Refine a research question, conduct systematic literature searches, meta‑analysis planning.	`Use $academic-research-suite to build a systematic review protocol for AI in higher‑education QA.`
academic‑paper	Draft outlines, abstracts, sections, format citations, add AI‑disclosure statements.	`Use $academic-research-suite to turn these notes into an IMRaD paper outline.`
academic‑paper‑reviewer	Simulated peer‑review, generate decision letters, suggest revisions.	`Use $academic-research-suite to review this manuscript and produce a journal‑style decision letter.`
academic‑pipeline	End‑to‑end “research‑to‑paper” process with integrity gates, revision loops, and final checks.	`Use $academic-research-suite to run an end‑to‑end pipeline from topic to revised manuscript.`
sr‑screener	Title/abstract or full‑text screening against a pre‑defined protocol; produces audit reports.	`Use $academic-research-suite: sr‑screener ta‑screen these records against my confirmed protocol.`
experiment‑agent	Plan code experiments, define statistical analyses, check reproducibility.	`Use $academic-research-suite to plan a code experiment and define reproducibility checks.`

Each workflow is invoked through the single skill entry; the router decides which internal mode to run based on the wording of your request.

Typical interaction pattern

State the goal and current assets – e.g.

Use $academic-research-suite.
Goal: write a journal article on AI adoption in higher education.
Current materials: literature matrix, rough findings, no outline.
Constraints: English, APA 7, policy‑maker audience.

Let the router pick the appropriate workflow (often deep‑research first if the question is vague).
Follow up with the specific mode – e.g., after the question is narrowed, ask for an outline via ars‑outline or academic‑paper.
Use the built‑in integrity gates – the pipeline can automatically check citation validity, flag unverified statements, and enforce reviewer‑style checks.

What it does not do

It does not train or fine‑tune any model.
It does not provide a standalone web UI – interaction happens inside a Codex conversation or via the CLI codex exec sandbox.
The optional full‑runtime adapter (agent‑team topology, hooks) is disabled by default; normal use relies on Codex’s native sub‑agent handling.

Who might find this useful?

Graduate students or researchers who want a structured, AI‑assisted way to manage literature reviews, drafting, and peer‑review simulations without building their own prompt library.
Research labs that already use Codex and need a plug‑and‑play skill for academic writing pipelines.
Tool developers looking for a reference implementation of a multi‑step research workflow that can be adapted to other LLM platforms.

Quick start checklist

Install the plugin (Marketplace or manual).
Open a fresh Codex chat.
Run /skills – confirm the single ARS entry.

Issue a request like:

Use $academic-research-suite.
I want to write a paper on AI governance in universities. I don’t have a clear question yet.

Follow the Socratic narrowing questions, then move to ars‑outline or academic‑pipeline as needed.

All details above are taken directly from the repository’s README; no additional features have been inferred.