The Web Search Verifier is a specialized verification framework designed for AI agents tasked with gathering high-stakes financial data such as interest rates, stock indices, and exchange rates. Instead of relying on single-source information, this skill enforces a mandatory triple-source verification protocol, requiring agents to cite original text snippets and match values across independent platforms like FRED, Bloomberg, and Trading Economics. It serves as a primary defense against data hallucinations by mandating strict source hierarchy and providing structured error handling for data discrepancies, ensuring that every data point is accurate, fresh, and verifiable.
主要功能
01Structured JSON output schema for automated downstream data processing
02Strict variance checking with a ±1% tolerance for data consistency
03Mandatory original text citation for evidence-based reporting and auditing
049 GitHub stars
05Triple-source cross-verification protocol requiring 3+ independent sources
06Tier-based source allowlist prioritizing official central bank and exchange data
