Manipulation Index 作成者: TrustInTheProcess

Manipulation Index (MX) analyzes web pages and shows a simple gauge meter in the corner of every page, indicating the content's emotional tone at a glance.

It combines three layers of analysis:

• Word-level sentiment — positive and negative vocabulary
• Tone signals — emotional patterns (hostility, warmth, alarm, optimism, etc.)
• Manipulation detection — fear framing, urgency pressure, divisive rhetoric

WHAT THE GAUGE SHOWS:

Green — Positive: uplifting, encouraging, warm content
Gray — Neutral: factual, balanced reporting
Red — Negative: critical, hostile, or distressing content
Warning — Manipulative: uses fear, division, or urgency tactics

WHAT IT DETECTS:

• Fear framing — "They don't want you to know...", doomsday scenarios
• Urgency tactics — "Act now before it's too late!", countdown pressure
• Divisive language — "Us vs. them", "enemies of the people"
• Emotional push — Outrage bait, indignation triggers

WHAT IT DOES NOT DETECT:

• Political bias (left vs. right)
• Factual accuracy (true vs. false)
• Source credibility

MX analyzes HOW something is said, not WHAT is being said.

FEATURES:

• Click the gauge for a detailed donut chart showing tone signal breakdown
• Click any category to see the exact words and phrases that were matched
• Full MX Report with word-level evidence for every score
• Drag the indicator anywhere on the page
• Settings: indicator size, opacity, position, and per-site disable

PRIVACY:

Everything runs locally in your browser. No data is collected, no external requests are made, no tracking of any kind.