A label on a headline is only as good as the search behind it, and what makes this usable is that the supporting and contradicting evidence comes back alongside the label — so the claim can be checked rather than taken. Two limits worth knowing before relying on it: every call depends on Gemini being available and on your quota, and the project notes its RSS parsing is simplified, so treat trending topics as a rough feed rather than a ranked index.
A small Python MCP server with two tools: one runs a headline through Google Gemini together with a web search and returns a verdict with its evidence, the other lists trending news topics from RSS feeds.
- A verdict on a headline — TRUE, FALSE, PARTIALLY_TRUE, UNVERIFIED or MISLEADING — with a confidence score, a truthfulness percentage, the supporting and contradicting evidence found, and recommendations — `fact_check_headline`
- Current trending topics with titles, sources and URLs, scoped to international, local or India — `get_trending_topics`
Python 3.11 or higher and a Google Gemini API key in `GEMINI_API_KEY`; nothing works without it. Install is a clone plus `pip install -r requirements.txt`, then `python -m factchck.news_factcheck` — which is also the client command, with the key in the entry's env block. `NEWS_API_KEY` and `SEARCH_API_KEY` are optional and widen the search. The project is `factchck` 3.0.0. It needs an internet connection on every call, and the checks run against your own Gemini quota.
Build from source — clone the repository and build it, then point your client at the binary