News & Sentiment

News intelligence over free public feeds: headline search (GDELT + RSS), full article extraction, aggregate sentiment, and per-day entity timelines. Built for brand monitoring, press tracking, and market-mood analysis.

What it scrapes

Source	Data
GDELT + RSS feeds	Headlines matching a query, with source, published date, and URL
Pry scraper (article extraction)	Full article text, summary, and metadata from a URL
GDELT tone analysis	Sentiment aggregates — avg tone, positive/negative percentage, timeline

Endpoints

Router: routers/news_data.py · tag: News Data

Method	Path	Summary
GET	/v1/news/headlines	Recent headlines matching a query (timeframe e.g. 24h, 7d)
GET	/v1/news/article	Extract a full article by URL (text + summary + metadata)
GET	/v1/news/sentiment	Aggregate sentiment for a query (tone + keyword fallback)
GET	/v1/news/timeline	Per-day entity timeline with tone and top sources

GET /v1/news/headlines

Query params: query (required, e.g. crypto regulation), limit (1–250, default 20), timeframe (GDELT timespan, e.g. 24h, 7d).

curl -X GET "http://localhost:8005/v1/news/headlines?query=crypto%20regulation&limit=5" \
  -H "Authorization: Bearer <key>"

Response (200):

{
  "success": true,
  "query": "crypto regulation",
  "limit": 5,
  "timeframe": "24h",
  "data": [
    {
      "title": "New EU rules target unbacked crypto assets",
      "source": "Reuters",
      "published_at": "2026-08-16T09:30:00Z",
      "url": "https://example.com/articles/eu-crypto-rules"
    }
  ]
}

GET /v1/news/article

Query params: url (required, article URL).

curl -X GET "http://localhost:8005/v1/news/article?url=https%3A%2F%2Fexample.com%2Farticles%2Feu-crypto-rules" \
  -H "Authorization: Bearer <key>"

Response (200):

{
  "success": true,
  "url": "https://example.com/articles/eu-crypto-rules",
  "data": {
    "title": "New EU rules target unbacked crypto assets",
    "text": "European regulators published draft rules on Monday...",
    "summary": "EU publishes draft rules for unbacked crypto assets.",
    "published_at": "2026-08-16T09:30:00Z",
    "author": "Jane Doe",
    "source": "example.com"
  }
}

GET /v1/news/sentiment

Query params: query (required), timeframe (e.g. 24h, 7d), limit (1–250, default 20).

curl -X GET "http://localhost:8005/v1/news/sentiment?query=Bitcoin%20ETF&timeframe=7d" \
  -H "Authorization: Bearer <key>"

Response (200):

{
  "success": true,
  "query": "Bitcoin ETF",
  "timeframe": "7d",
  "data": {
    "avg_tone": 2.4,
    "positive_pct": 58.0,
    "negative_pct": 21.0,
    "neutral_pct": 21.0,
    "articles_analyzed": 120,
    "timeline": [
      {"date": "2026-08-15", "avg_tone": 1.8, "articles": 18}
    ]
  }
}

GET /v1/news/timeline

Query params: entity (required, e.g. OpenAI), timeframe, limit.

curl -X GET "http://localhost:8005/v1/news/timeline?entity=OpenAI&timeframe=7d" \
  -H "Authorization: Bearer <key>"

Response (200):

{
  "success": true,
  "entity": "OpenAI",
  "data": [
    {
      "date": "2026-08-16",
      "avg_tone": -0.6,
      "article_count": 42,
      "top_sources": ["TechCrunch", "The Verge", "Bloomberg"]
    }
  ]
}

MCP tools

The Pry MCP server exposes the news lane as two tools (71 tools total):

Tool	Backing endpoint
pry_news	GET /v1/news/headlines
pry_sentiment	GET /v1/news/sentiment

{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "pry_news",
    "arguments": {"query": "crypto regulation", "language": "en", "limit": 5}
  }
}

Both tools are x402 pay-per-call at $0.002/call (see /v1/x402/pricing).

Who uses this

• Brand monitoring — track press mentions of a company or product.
• Market-mood analysis — pry_sentiment before a trade or campaign.
• Research pipelines — headline + article extraction for dataset building.
• Competitor tracking — entity timelines for narrative shifts.

Next steps

• App Store — track app press alongside store data
• MCP Integration — run the tools from Claude / Cursor / Hermes