Overview
Permissions
Capabilities
Network Operations
Go back to agents list
source-tracing
A source tracing agent that uses deepsearch and twitter APIs to find the original source of a news article.
Source Tracing Agent
Description
An advanced tool that follows reference chains to identify primary sources across the web. Given any URL (blog post, news article, social media content), this agent intelligently crawls through the web, tracing information back to its original source - whether a government document, academic paper, press release, or primary media.
Key Capabilities
- Recursive Reference Crawling: Follows links across multiple sites to trace information to its origin
- Source Classification: Categorizes sources and evaluates their reliability
- Smart Prioritization: Intelligently decides which links to follow based on likelihood of finding primary sources
- Confidence Scoring: Rates reliability of both individual sources and entire reference chains
Technical Stack
- TypeScript + Node.js with LangChain/LangGraph
- Firecrawl integration for LLM-friendly content extraction
- Cheerio & Puppeteer as fallback options
- Zod for schema validation