# AI use cases for Absent Author (THUROI0787/absent-author)

Repo: [THUROI0787/absent-author](https://github.com/THUROI0787/absent-author) | Use cases: 4 | Page: https://dedupedai.com/use-cases/thuroi0787-absent-author/

An evidence list and two agent skills for spotting research papers that AI produced and no human stood behind, plus Python checks that flag candidate signs for a person to read.

This file holds every use case of this repo. Each one starts with a "## " heading and ends with its URL.

- Check an AI-assisted paper before you sign it: https://dedupedai.com/use-cases/thuroi0787-absent-author/author-pre-submission-pass/
- Grade a suspicious submission with evidence, not guesses: https://dedupedai.com/use-cases/thuroi0787-absent-author/reviewer-triage-suspicious-submission/
- Scan a paper folder for AI-style surface traces: https://dedupedai.com/use-cases/thuroi0787-absent-author/lint-paper-for-surface-traces/
- Check references, repeated numbers, and hidden PDF text: https://dedupedai.com/use-cases/thuroi0787-absent-author/verify-references-numbers-and-pdf/

## How to check an AI-assisted paper before you sign it with Absent Author

Repo: [THUROI0787/absent-author](https://github.com/THUROI0787/absent-author) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README

Run paper-author-pass to get questions only you can answer, a claim-evidence map, and verified references.

**Who it is for:** An author finishing an AI-assisted paper who must own every claim before submission.

**The problem:** AI-drafted sections can carry unverified references, numbers that disagree between text and tables, and claims no one checked.

**The outcome:** A list of decision questions for you, a claim-evidence map, a change log, and three questions to answer before you sign.

### Steps

1. Clone the repo with `git clone https://github.com/THUROI0787/absent-author.git && cd absent-author`.
2. Install the skills with `./install.sh`.
3. Open Claude Code or Codex in your paper folder and tell the agent to use paper-author-pass on paper/ in audit mode.
4. Answer the questions the skill asks; it does not answer them for you.
5. Read the claim-evidence map and fix any claim without evidence before you sign.

### Prompt to paste into your agent

```text
Use the absent-author repo's paper-author-pass skill on <paper folder> in audit mode. Ask me the questions only the author can answer, then produce a claim-evidence map and a change log. Verify that every reference the skill marks as verified can be found by you in the bibliography.
```

### Caveat

The skill asks questions and gives you a checklist; it cannot confirm that the research itself is sound, and it should not be used to answer its own questions.

Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/

URL: https://dedupedai.com/use-cases/thuroi0787-absent-author/author-pre-submission-pass/

## How to grade a suspicious submission with evidence, not guesses with Absent Author

Repo: [THUROI0787/absent-author](https://github.com/THUROI0787/absent-author) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README

Run paper-slop-screen to get W/R/Q grades, a quadrant, located findings, and a note to the AC.

**Who it is for:** A reviewer or area chair facing a submission that looks like it was produced end to end by an agent with no human owner.

**The problem:** Reviewers have intuitions about AI-produced papers but lack a shared, citable vocabulary with evidence behind it.

**The outcome:** W/R/Q grades, a quadrant placement, every finding with a location and quote, a substance-only review paragraph, and a note to the AC.

### Steps

1. Check your venue's reviewer LLM policy before you run the screen on a submission under review, as the README asks.
2. Clone the repo with `git clone https://github.com/THUROI0787/absent-author.git && cd absent-author`.
3. Install the skills with `./install.sh --project` or `./install.sh` for your user folder. (inferred, check the README)
4. In Claude Code or Codex, tell the agent to use paper-slop-screen to triage the paper.
5. Check every finding against its location and quote before you cite it in a review.

### Prompt to paste into your agent

```text
Use the absent-author repo's paper-slop-screen skill to triage <paper PDF or folder>. Return the W/R/Q grades, the quadrant, and each finding with its location and quote. Verify one finding by opening the cited location yourself before you rely on it.
```

### Caveat

The README says language evidence never raises the research grade and the output is not a verdict on AI use, so the screen supports a review rather than replacing one.

Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/

URL: https://dedupedai.com/use-cases/thuroi0787-absent-author/reviewer-triage-suspicious-submission/

## How to scan a paper folder for AI-style surface traces with Absent Author

Repo: [THUROI0787/absent-author](https://github.com/THUROI0787/absent-author) | Difficulty: Intermediate | Time: 15 minutes | Basis: From the README

Run slop_lint.py for counts of 28 surface traces, each with a file, line, and snippet.

**Who it is for:** Anyone who wants a quick first pass over a draft or preprint before a deeper read.

**The problem:** Reading a full paper to spot writing patterns takes time, and the patterns are easy to miss on a first read.

**The outcome:** A markdown report of calibrated counts per check, with a band against human percentiles and a snippet for each hit.

### Steps

1. Clone the repo with `git clone https://github.com/THUROI0787/absent-author.git && cd absent-author`.
2. Run `python tools/slop_lint.py paper/ --source -o lint.md` on your paper folder.
3. Open lint.md and read each hit at its file and line.
4. Treat each hit as a candidate to judge, not a verdict.

### Prompt to paste into your agent

```text
In the absent-author repo, run tools/slop_lint.py on <paper folder> with --source and write the output to lint.md. Summarize the checks in the high band and list their file, line, and snippet. Verify by reading two of the cited lines yourself.
```

### Caveat

The README says a clean report proves nothing, and the calibration sample is small, so the counts only flag writing patterns for a human to read.

Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/

URL: https://dedupedai.com/use-cases/thuroi0787-absent-author/lint-paper-for-surface-traces/

## How to check references, repeated numbers, and hidden PDF text with Absent Author

Repo: [THUROI0787/absent-author](https://github.com/THUROI0787/absent-author) | Difficulty: Intermediate | Time: 30 minutes | Basis: From the README

Run three helper scripts that report candidates for references, numbers, and hidden text.

**Who it is for:** An author or reviewer with a bibliography, a LaTeX source, and a PDF who needs the slow verification work done.

**The problem:** Checking every author in every reference and every restated number by hand is slow, and hidden or remapped PDF text is hard to see.

**The outcome:** Candidate lists for reference author checks, numbers stated with different values, and hidden or remapped text, each to be checked by a person.

### Steps

1. Clone the repo with `git clone https://github.com/THUROI0787/absent-author.git && cd absent-author`.
2. Run `python tools/ref_verify.py refs.bib --sample 5 --mailto you@example.org` on your bibliography.
3. Run `python tools/number_ledger.py paper/main.tex` on your LaTeX source.
4. Run `pip install pymupdf` and then `python tools/pdf_hidden_text.py paper.pdf` on the PDF.
5. Open each candidate at its location and decide whether it is a real problem.

### Prompt to paste into your agent

```text
In the absent-author repo, run tools/ref_verify.py on <refs.bib>, tools/number_ledger.py on <main.tex>, and tools/pdf_hidden_text.py on <paper.pdf>. List every candidate with its source location. Verify each candidate by reading the cited source yourself, since the tools report candidates, not verdicts.
```

### Caveat

The helpers report candidates for a person to read and never verdicts, and the PDF check's OCR step needs optional system tools that the README lists separately.

Disclaimer: Third-party repos we don't own or vouch for. AI-generated use cases can be wrong: review everything and run at your own risk. Full text: https://dedupedai.com/disclaimer/

URL: https://dedupedai.com/use-cases/thuroi0787-absent-author/verify-references-numbers-and-pdf/