Skip to content
SkillModel-led4,645 tokens in use

dbs-ai-check

Scans long articles, video scripts and posts for 22 signs of AI writing, quotes each hit with a severity, and rewrites only after asking what you meant.

11,518 stars on the repository Custom / Unknowndontbesilent2025/dbskill · skills/dbs-ai-check
npx -y skills add dontbesilent2025/dbskill -g --all
View source

Install command as written in the repository's README

Skill label

measured from its files

What installing dbs-ai-check costs

114tokens idle
4,645tokens in use

Idle is what sits in context while the skill is installed. In use is its instructions, loaded when it runs. Counts are approximate, made offline with cl100k_base.

Model dependence Model-led

Script-ledMixedModel-led

Mostly instructions, so how well it works depends on the model that follows them.

This describes the skill's structure, how much of it is scripts. It does not say how well it works.

Touches

Nothing detected: no shell, network, file writes, git or credentials in its files.

Declared tags come from the skill's frontmatter; detected ones from its scripts, its code blocks and the commands it names in inline code. This is detected, not audited.

How it scores

Quality

How well the skill is written, counted from its own files: its frontmatter, its steps, examples, stated output and constraints, its size and whether every file it ships is one it names.How we score it →
Fair
  • Frontmatter8/15
  • Structure16/25
  • Examples9/15
  • Output0/15
  • Size15/15
  • Resources0/15

Security

Starts at 100 and loses points for each thing in the skill's files that a reader should look at first.How we score it →
Excellent

No findings.

Repository health

Pre-computed score based on development activity, maintenance, community, maturity, and trend momentum.How we score it →
Excellent
  • Development activity96
  • Maintenance100
  • Community64
  • Maturity28
  • Momentum40

dbs-ai-check is a detector for AI writing traces in Chinese-language copy such as official account articles, short video scripts and tweets. It reads the text in order, quotes every passage that matches one of 22 listed patterns, explains the problem in a sentence and tags it with a pattern number and a severity.

By default it only reports. If you then say you want to revise, it goes through the hits from most to least severe, asks one question about what you were trying to achieve, and gives a direction to write from, never a finished rewrite. It argues that removing AI traces does not make content better.

What You Get

  • A report that lists each hit in text order with the original passage quoted
  • A pattern number and a severity (strong, medium or weak signal) for each hit, plus a one or two sentence summary
  • A note on the genre it assumed when you did not give one
  • A direct statement when the text shows little AI flavor, with no forced findings
  • A guided revision round on request, one question per hit, ending in a direction for you to write from

Common Use Cases

  • Checking a long article draft before publishing
  • Reviewing a short video script, where closing punchlines on every paragraph are normal for the format
  • Finding which paragraphs of a tweet thread read as machine-made
  • Getting editor-style questions about the intent behind a repeated sentence structure

What Makes It Unique

Most humanizer tools rewrite text to hide AI traces. This one states that the goal is to help you see your own writing, so it diagnoses by default and refuses to rewrite before you answer a question about what you meant. It also lists false-positive warnings, such as contrast phrases that classic writers use, and adjusts the rules by genre.

Under The Hood

This Skill is instructions only. The agent reads SKILL.md, which holds the 22 patterns, the report format, the question for each pattern and the genre rules, and applies them to the text you give it. It has no scripts, runs no commands, makes no network calls and writes no files.

Join founders building with open source

Opinionated takes, migration guides, cost-saving tips, and insights from the open source ecosystem.

Subscribe on SubstackJoin 750+ subscribers