ai-text-detector-github
Skill549 starsChanged 6 months agoMITView on GitHub
Detect whether a passage shows AI-like writing signals and return an explainable risk estimate with confidence, caveats, and next steps.
In this repository2 other files it gives its agents.
- AGENTS.mdAGENTS.md
- ai-detectorSkill
The file
We split this skill into the rules it gives an agent
10 rules. About 0 are common, 10 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈10
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 25
Use this skill when:
- Distinctive: line 34
Never present the score as proof.
- Distinctive: line 35
Never accuse a named person of cheating, fraud, or misconduct.
- Distinctive: line 36
Ask for a longer sample when the text is under about 80 words.
- Distinctive: line 37
Prefer "AI-like signals are present" over "This was written by AI".
- Distinctive: line 43
If the sample is under about 80 words, explain that the detector will be noisy and ask for a longer sample when possible.
- Distinctive: line 45
Run the local detector:
- Distinctive: line 104
Keep the analyzer explainable and lightweight.
- Distinctive: line 105
Prefer local heuristics over hidden network calls.
- Distinctive: line 106
Update this file whenever behavior or output changes.
≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
AI Text Detector Skill
Overview
This skill wraps the local ai-detector-skill analyzer into a reusable open source Codex-style skill.
It is designed for:
- essays
- emails
- articles
- reviews
- forum posts
- other prose where a user asks whether the writing may be AI-generated
This skill returns a risk estimate, not proof of authorship.
Trigger Conditions
Use this skill when:
- the user asks "is this AI-written?"
- the user asks to detect AI-generated text
- the user pastes a passage and asks whether it sounds machine-written
- the user wants a cautious AI-likeness review for a document or message
Required Behavior
- Never present the score as proof.
- Never accuse a named person of cheating, fraud, or misconduct.
- Ask for a longer sample when the text is under about 80 words.
- Prefer "AI-like signals are present" over "This was written by AI".
- For high-stakes contexts, recommend human review and comparison with known writing samples.
Execution Flow
- Receive the candidate text and check length.
- If the sample is under about 80 words, explain that the detector will be noisy and ask for a longer sample when possible.
- Save the text to a file or pipe it through stdin.
- Run the local detector:
ai-detect path/to/text.txt --jsonor:
python -m aidetect.cli path/to/text.txt --jsonor from the repository helper:
python scripts/detect.py path/to/text.txt --json- Parse the JSON result and validate that it includes:
scoreconfidenceverdictconclusionsignalscaveats
- Respond in this order:
- One-sentence conclusion with uncertainty.
- Score and confidence.
- Strongest evidence signals.
- Caveats.
- Next steps only when useful.
Output Guidance
Good phrasing:
- "AI-like signals are present, but this is not proof."
- "The result is uncertain because the sample is short."
- "This should be reviewed against known writing samples."
Avoid:
- "This was definitely written by AI."
- "The detector proves misconduct."
- Any accusation against a named person.
Repository Layout
SKILL.md: skill contract and usage rulesscripts/detect.py: repository-local wrapper for the detectorscripts/setup.sh: local environment bootstrapreferences/api-reference.md: response contract and CLI referenceassets/templates/report.md: reusable response template
Development Notes
- Keep the analyzer explainable and lightweight.
- Prefer local heuristics over hidden network calls.
- Update this file whenever behavior or output changes.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

