Absent Author skill
An evidence list of papers made by AI with no human behind them and two skills: a pre-submission pass and a review of a manuscript
- 148 ★ on the repository
- Last commit
- MIT
Minimal access
Why this level
- Reference checks query external databases and send titles and manuscript data to them
- The result is easy to read as an accusation, so conclusions about people are drawn only by hand and with sources
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/absent-authorDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewMinimal accessThe tool runs nothing and makes no network calls.
- Reference checks query external databases and send titles and manuscript data to them
- The result is easy to read as an accusation, so conclusions about people are drawn only by hand and with sources
This is third-party code. Review the repository files before installing.
What it does
Absent Author is a repository by researchers from several universities. It collects signs of research papers that AI produced end to end with no human who steered or checked the work. Every sign is backed by sources, and the tools turn them into evidence that a reviewer can cite and an author can fix. Authors get the `paper-author-pass` skill: it asks questions only a human can answer, builds a claim to evidence map, checks references and numbers, then looks at style and ends with three questions. Reviewers and editors need the second skill to examine someone else manuscript. Version 0.5.0 added checks of the full author list in references, number consistency and hidden prompt injections in the paper text. The README comes in English, Chinese and Japanese, with a source summary in `docs/SOURCES.md` and an archived copy on Zenodo.
Who it is for. Authors of research papers who finish a text with AI help, and reviewers and journal or conference editors checking manuscripts.
Good fit when
- A paper was written with AI help and before submission you need to see what needs your own input
- A reviewer needs concrete evidence in place of a general feeling that the text is AI-made
- References and numbers in a manuscript have to be checked for consistency
- The text is suspected of hidden instructions aimed at reviewer models
Not a fit when
- You need an AI text detector with a probability percentage: the skill gathers evidence and issues no verdict
- You need to prove authorship in a dispute or accuse a specific person
- The text is not a research paper: the signs are tuned for peer-reviewed publications
Example request
Run my manuscript through paper-author-pass, show weak spots in references and numbers and ask the three questions I must answerLimitations
The skill collects observable signs, but the final decision belongs to a human. Reference checks depend on external databases, and hidden model instructions are sometimes placed by a venue as a canary, so version 0.5.0 separates such canaries from author misconduct. The last commit is dated November 4, 2026. The authors state that the work is not a legal opinion.
How to disable. Delete the `paper-author-pass` skill folder and the second skill from the agent skills directory (`.claude/skills` or `.agents/skills`) where you copied them.
FAQ
Does the skill rule that a paper was written by AI?
No, it gathers signs with sources and shows them. The conclusion stays with the author, reviewer or editor.
Who is paper-author-pass for?
For an author finishing a paper with AI help: the skill asks questions only a human can answer and checks references and numbers.
What is new in version 0.5.0?
A full author list check in references, number consistency and hidden model instruction detection. Venue canaries are separated from misconduct.
Which languages is the documentation in?
The README is available in English, Chinese and Japanese.
Where is the source summary?
In `docs/SOURCES.md`, which gathers links to the material the signs are built on.
Discussion
No questions or reviews yet.
Related
SkillsA skill set for academic work: paper review, pre-analysis plan check, grant proposal review, code audit and slide decks.
42 skills for the whole research paper path: from discovery and drafting to citation checks, submission, rebuttal and publication
paper-reading-skill turns one arXiv AI or robotics paper into a report with each claim tied to evidence in the source.
AI Vibe Writing Skills
AI-Vibe-Writing-Skills
Skill pack for academic papers covering author voice, evidence-based arguments, sentence flow and formal replies to reviewers.
A skill for researchers who draft in Chinese: drafts and results become an English manuscript, with citations and claims checked
A skill set for academic writing: source notes, citation audits, claim checks and local deterministic validation.
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Authentication | Repository stars | Main difference |
|---|---|---|---|---|
| Absent Author (this card) | no | 148 | ||
| Research Paper Writing | no | 8,329 | checks a manuscript and looks for signs of an absent author, with no writing | |
| Paper Search MCP | no | key is optional | 3,076 | works with a finished manuscript and does not search paper databases |
README in short
The README presents the project as an evidence list plus two skills for spotting papers that have no human author. It continues with release news, a table of roles (author, reviewer) with the matching tool and what each gets. Links lead to sources and a changelog.