financial news event verifier

The problem

Financial news feeds are often cluttered with stale reports, duplicate headlines, and rumors disguised as facts. Relying on unverified claims for market analysis leads to poor decision-making based on misinterpreted or outdated events.

What it does

• Identifies and flags exact normalized duplicates and stale original reports within news datasets.
• Validates claims against primary filings or official statements to distinguish between rumors, proposals, and completed events.
• Inspects metadata for future-dated publication times and extraction uncertainty in OCR or transcripts.
• Generates a structured claim packet containing evidence excerpts, chronology, and unresolved contradictions.

Frameworks & tools

The skill uses a Python-based analysis script and requires data to be mapped to a specific JSON input contract. It is designed for integration with custom data pipelines and human-in-the-loop review systems.

Why this beats prompting it yourself

General-purpose prompts often miss the nuance between an event and an interpretation. This skill enforces a rigorous verification workflow that treats metadata as first-class evidence, preventing the hallucination of facts from rumors or misinterpreted sentiment.

Use cases

• Auditing high-frequency news feeds for data integrity before they hit a research database.
• Cross-referencing press releases against official regulatory filings to confirm event completion.
• Scrubbing historical news datasets of duplicates and future-dated entries for backtesting.

Known limitations

Does not include a browser, OCR, or media forensics tools. Requires users to map external data to the provided JSON contract manually or via a separate script.