ContentLens – Settings

General

Default Detection Threshold 7

Content scoring ≥ this value will be filtered. Lower = more aggressive filtering.

1–3 · Human-like

4–6 · Ambiguous

7–10 · AI-like

Default Filter Mode

Blur — blurs suspected AI content. Hover to reveal. Hide — removes flagged content from view entirely. Highlight — marks content with a red outline (non-intrusive). Badge — shows a score badge only. Content is unmodified.

Minimum Text Length 80 chars

Text blocks shorter than this will be skipped.

Domain Rules

Disable ContentLens on specific websites (e.g., banking, healthcare, internal tools).

Enter a hostname to disable ContentLens on.

No disabled domains yet.

Privacy

ContentLens is 100% local.

All scoring is done on-device using heuristics.
No page content is ever sent to a server.
Only text hashes and scores are stored (no raw text).
No user identity or browsing history is collected.

Cache Management

Cached scores: —

Clearing the cache will cause all pages to be re-scored on next visit.

About ContentLens

ContentLens estimates the likelihood that text content on web pages was generated by an AI system. It uses local statistical heuristics — no internet connection required.

⚠ Scores represent probabilistic similarity, not proof of AI origin. ContentLens does not claim definitive AI detection. Always use your own judgment.

Version 1.0.0 Manifest V3 Local-only scoring