# Review Text for Possible AI Writing Patterns

Writing patterns alone cannot reliably establish whether content was generated by AI. This skill organizes a qualitative assessment into an HTML report with supporting observations.

## Install

```bash
npx skillstore add notegpt-official/ai-detector
```

## Metadata

- Status: approved
- Slug: notegpt-official-ai-detector
- Skillstore revision: r1
- Version status: missing
- Tree hash: 4cbc6705519e0d36c8de70de28750e13698be6a745058e8d55d033d5fffdcdb9
- Author: notegpt-official
- GitHub username: notegpt-official
- Repository: https://github.com/notegpt-official/notegpt-skills/tree/bd26f364942d87681cb0d2dbfd0e21288a5c5298/skills/ai-detector
- Ref: c993c6abcbffafea2683d7748b53b3dfaf438415
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: allowed
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Quality score: 72
- Public page: https://skillstore.pages.dev/skills/notegpt-official-ai-detector
- Manifest: https://skillstore.pages.dev/api/skills/notegpt-official-ai-detector/manifest

## Capabilities

- Review user-provided content for patterns associated with AI generation.
- Assess the apparent likelihood of AI generation.
- Generate an HTML report with evidence supporting the assessment.
- Represent the likelihood score with a percentage-based circular indicator, ring fill, and color.
- Include a download link for the analyzed content in the report.

## Use Cases

- Editorial Content Review: Review a submitted article for possible AI writing patterns before discussing specific observations with its author.
- Teaching Discussion Preparation: Prepare an exploratory review of a consented writing sample without treating the result as evidence of academic misconduct.
- Draft Self-Assessment: Examine your own draft for repetitive or generic patterns and review the supporting observations in an HTML report.

## Prompt Templates

### Review a Short Passage

```
Review this text for possible AI-generated patterns: [text]. Produce an HTML report and clearly state that the assessment cannot prove authorship.
```

### Request Supporting Evidence

```
Assess this draft: [text]. In the HTML report, quote relevant passages, explain each observation, and identify plausible human explanations.
```

### Specify the Report Elements

```
Analyze [text] and generate an HTML report. Include a circular percentage indicator with matching fill and color, evidence, limitations, and a content download link.
```

### Account for Context and Uncertainty

```
Review [text] given this context: [genre, editing history, language background]. Separate observations from inference, explain uncertainty, and label any percentage as uncalibrated.
```

## Limitations

- The skill provides instructions, not a validated detection model or calibrated probability estimator.
- Writing patterns can produce false positives and false negatives; results cannot prove authorship.
- The source provides no accuracy benchmarks, language coverage guarantees, or scoring methodology.
- HTML rendering and download behavior depend on the host environment; no implementation or browser testing is included.

## Best Practices

- Use the report as a discussion aid and verify concerns through independent evidence.
- Provide writing context and request alternative explanations for each observed pattern.
- Remove sensitive information before submission and inspect the report and download link before sharing.

## Anti Patterns

- Treating a generated percentage as a measured probability or proof of AI authorship.
- Making disciplinary, hiring, or publication decisions solely from this assessment.
- Submitting confidential content without permission or reviewing the host service's data handling.

## Security Audit

- Audited at: 2026-11-06T15:33:15.682\+00:00
- Summary: Static analysis reported no findings, so no finding verdicts were required. The prompt requests content analysis and HTML reporting; no evidence found of prompt injection, credential access, or data exfiltration. Its likelihood scores have no documented validation and should not establish authorship.

## Stats

- Views: 0
- Downloads: 0
- Favorites: 0
- Popularity score: 0