技能 人工智能 AI 生成文本检测

AI 生成文本检测

v20261014
detect-ai-text
调用外部检测 API 判断文档正文是否由 AI 撰写,输出未校准的概率值、具体语言特征,并对非散文内容诚实弃权。适用于作文、简历、投稿与文章的可疑性初筛,仅作参考信号,不能作为作者身份或学术不端的判定依据。
获取技能
337 次下载
概览

AI-Text Detection

Inspect a provider’s uncalibrated AI-style signal for a document’s prose, with the specific linguistic tells and an honest abstention when the document isn't prose. Uses an external Stipple API; verify current access, cost and privacy terms.

When to use

  • Educators screening student submissions
  • Publishers and platforms triaging inbound content
  • HR reviewing AI-drafted CVs (flag, don't reject)
  • Researchers checking source provenance

Instructions

  1. Get the document. URL, local file path (PDF, DOCX, TXT, MD), or raw text via --text.
  1. Run detection.

```bash
curl --fail --show-error --silent --max-time 30 -X POST https://www.stipple.sh/v1/detect-ai-text \
-F "file=@essay.pdf" \
-H "Authorization: Bearer $STIPPLEAPIKEY"
```

For raw text: POST JSON {"text": "..."} to the same endpoint.

  1. Interpret the response.
  • applicable: false — the document is not prose (forms, tables, scans, spreadsheets). Detection is deliberately refused rather than guessed. Report this and stop.
  • probability — model confidence (0–1), NOT a calibrated truth
  • lean — "ai" | "human" | "unsure"
  • tells[] — the specific phrases/patterns flagged (e.g. "It is important to note that", uniform sentence length, low burstiness)
  • reasoning — why the model reached its verdict
  • limitations — always present; the API states its own noise profile
  1. Report honestly. This measures style, not authenticity:
QuestionTool
Was this written by AI? (style)this skill
Is this document genuine/tampered? (forensics)verify-document skill

A human can write generically; an AI can write plainly. One triage signal, never a verdict.

Output format

Provider AI-style score: 0.87 (uncalibrated; lean: ai; not authorship proof)
prose ratio: 0.82

linguistic tells:
  - "It is important to note that, in today's fast-paced world" (stock phrase)
  - uniform sentence length across paragraphs
  - low burstiness; no authorial asides

reasoning: The text relies entirely on formulaic transition clichés...
limitations: The probability is the model's CONFIDENCE, not a calibrated truth.

Limitations and Safety

  • AI-text detection is probabilistic and can disproportionately flag templated,

coached, translated, accessibility-assisted, or non-native-English writing. Do
not use it as proof of misconduct or as the sole basis for employment, academic,
publishing, or disciplinary action.

  • Uploading a document sends its contents to a hosted third party. Obtain explicit

approval first, remove unnecessary personal or confidential material, and verify
the provider's current retention and deletion terms.

  • Preserve the original work and give the affected person a meaningful human review

and appeal path before any consequential decision.

Notes

  • Non-prose documents (forms, tables, payslips) get applicable: false — the engine refuses rather than guessing
  • Can false-flag templated/coached or non-native-English writing — always present as one signal alongside human review
  • Verify current authentication/allowance on https://www.stipple.sh; do not guarantee a free anonymous tier or publish the submitted document.
信息
分类 人工智能
名称 detect-ai-text
版本 v20261014
大小 3.89KB
更新时间 2026-10-16
语言