Programmatic Access to Text Analysis for 67+ Manipulation Tactics
The Truth Seeker API allows developers to analyze text for manipulation tactics programmatically. Submit any text—news articles, social posts, speeches, marketing copy—and receive a detailed breakdown of detected maniptics.
Analyzes text for 67+ manipulation tactics and returns a comprehensive breakdown with scores, explanations, and warnings.
| Parameter | Type | Required | Description |
|---|---|---|---|
text |
string | Required | The text content to analyze. Maximum 10,000 characters. |
curl -X POST https://goms.life.repl.co/analyze \
-H "Content-Type: application/json" \
-d '{
"text": "Breaking: Experts agree this is the most dangerous threat we have ever faced. Everyone is talking about it. If you care about your children, you must act now!"
}'
const response = await fetch('https://goms.life.repl.co/analyze', {
method: 'POST',
headers: {
'Content-Type': 'application/json',
},
body: JSON.stringify({
text: 'Breaking: Experts agree this is the most dangerous threat we have ever faced...'
})
});
const result = await response.json();
console.log(result);
import requests
url = "https://goms.life.repl.co/analyze"
payload = {
"text": "Breaking: Experts agree this is the most dangerous threat..."
}
response = requests.post(url, json=payload)
result = response.json()
print(result)
{
"totalScore": 75.8,
"interpretation": "Heavy manipulation detected",
"constitutionalWarning": null,
"metrics": {
"tacticsTriggered": 8,
"totalInstances": 23,
"averageIntensity": 7.2,
"mlu": 85.4,
"mdq": 85.4,
"wordCount": 25,
"totalSeverity": 10.7
},
"breakdown": {
"fearMongering": {
"score": 8,
"count": 3,
"flaggedPhrases": ["most dangerous threat", "act now"],
"impact": "...",
"counterstrategy": "..."
},
"bandwagon": {
"score": 6,
"count": 2,
"flaggedPhrases": ["everyone is talking"],
"impact": "...",
"counterstrategy": "..."
}
}
}
| Field | Type | Description |
|---|---|---|
totalScore |
number | Overall manipulation score (0-100). Sophisticated multi-factor algorithm considering intent, breadth, severity, and volume. Higher = more manipulation detected. |
interpretation |
string | Human-readable interpretation of the score (e.g., "Minimal bias detected", "Heavy manipulation detected") |
constitutionalWarning |
object | null | Warning object if foundational attacks detected (Rights Inversion, Republic vs Democracy shifts, Consensus over Reason) |
metrics |
object | Detailed metrics object containing: |
tacticsTriggered |
number | Number of manipulation tactics detected |
totalInstances |
number | Total count of all flagged phrases/words across all tactics |
averageIntensity |
number | Average severity score of detected tactics (0-10 scale) |
mlu |
number | NEW! Manipulation Density Score (0-100). Capped version of MDQ for standardized comparisons |
mdq |
number | NEW! Manipulation Density Quantity (uncapped). Raw calculation showing true saturation level. Formula: (totalSeverity / wordCount) × 200. Can exceed 100 for extremely dense manipulation |
wordCount |
number | NEW! Total word count in analyzed text (used to calculate MLU) |
totalSeverity |
number | NEW! Sum of all tactic severity scores (0-10 each). Used to calculate MLU. Higher = more severe manipulation detected |
breakdown |
object | Detailed breakdown object with each detected tactic showing score, count, flagged phrases, impact, and counterstrategy |
totalScore (0-100): Sophisticated multi-factor analysis
MLU - Manipulation Density (0-100): Capped concentration metric
MDQ - Manipulation Density Quantity (uncapped): True saturation level
Use Cases:
When rate limit is exceeded, you'll receive a 429 status code with a Retry-After header indicating seconds to wait.
{
"detail": "Rate limit exceeded. Try again in 45 seconds."
}
| Status Code | Error | Description |
|---|---|---|
400 |
Bad Request | Missing or invalid text parameter |
422 |
Validation Error | Text exceeds 10,000 character limit |
429 |
Rate Limit Exceeded | Too many requests. Check Retry-After header |
500 |
Internal Server Error | Server error during analysis. Please retry |
The API currently detects 67+ manipulation tactics across these categories:
Build classroom applications that teach students to recognize manipulation in real-time
Analyze news feeds, social media posts, or press releases for manipulative content
Screen user-generated content for manipulation tactics before publication
Conduct large-scale studies on manipulation patterns in different media sources
Create plugins that analyze articles as users browse the web
Add manipulation detection to AI assistants and chatbots
Questions or feedback about the API? Found a bug or have a feature request?