Skip to content
Get packs

ARA Rigor Reviewer

Scores a research artifact's evidence, claims and methods on six dimensions and writes a severity-ranked review report.

Source Download skill
Repository
Orchestra-Research/AI-Research-SKILLs
Stars
14,195
Licence
MIT
Files
2 files

What it does

This skill teaches an AI agent to act as a strict peer reviewer for an Agent-Native Research Artifact (a structured folder of research claims, experiments, and logs). Once installed, the agent reads the artifact's files in a fixed order, checks whether each claim is actually supported by its cited experiments, whether falsification criteria are testable, and whether the research process is honestly documented, including failures. It then scores six dimensions from 1 to 5 and writes a constructive level2_report.json with strengths, weaknesses, and suggestions, ending in a recommendation from Strong Accept to Reject. The skill states it works only with the agent's own file tools and does not run code or fetch web pages.

Great for

  • Researchers preparing an Agent-Native Research Artifact for publication who want a critical pre-submission read
  • Lab leads or students who want a consistent, rubric-based critique of experiment design and claim scope
  • Developers building or testing ARA tooling who need a semantic review layer after structural validation

Skip it if

  • Judging whether research is correct in reality: the agent only reads the artifact and does not check external sources or rerun experiments, so its scores reflect textual reasoning, not verified truth
  • Replacing human peer review for high-stakes decisions such as grants, hiring, or publication acceptance, since the scores are model judgments and may be inconsistent
  • Artifacts that have not passed Level 1 structural validation, because the skill assumes structure is already sound

Try asking

  • “Review my Agent-Native Research Artifact and write a Level 2 report before I release it”
  • “Score the six review dimensions for this ARA and list the weakest claims with suggestions”
  • “Check whether each claim's falsification criteria are specific enough for someone else to test”
  • “Summarise the exploration tree and flag where failed attempts were left out of the record”
Step-by-step guide: how to use ARA Rigor ReviewerInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 2 original files, the MIT licence and an attribution note.

    rigor-reviewer.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip rigor-reviewer.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip rigor-reviewer.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Run Level 1 structural validation first, since this skill assumes it has already passed and does not recheck structure
  • Treat the scores as a starting point for discussion and verify any serious weakness yourself before acting on it
  • Ask for the report to focus on one dimension, such as falsifiability, if you want deeper feedback on that area

What's inside

  • references/review-dimensions.md
  • SKILL.md

Redistributed unmodified from Orchestra-Research/AI-Research-SKILLs under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.