I built an AI content agent that would rather drop a topic than make things up

Description and use

An AI agent produces fact‑based insurance articles weekly, halting when data is insufficient, with human approval before publishing.

Insurance sites need accurate, fact‑backed content; AI models often hallucinate, risking misinformation about legal obligations and coverage.

The agent runs weekly, producing accurate drafts that are reviewed and published by a human, earning a 5/5 client review.

Read Google Search Console and Analytics to identify content gaps
Select a topic from the site’s domain, ensuring no duplicate coverage
Research the topic using a fixed allow‑list of sources (slov-lex.sk, government bodies, regulator, insurers’ sites)
Apply a stop rule: if insufficient confirmed facts, drop the topic
Generate the article in Slovak with Claude using only collected facts
Run an independent AI review to detect factual errors, outdated data, invented references, and source consistency
Save the verified draft as a WordPress draft for human review
Human publishes or rejects the article via dashboard

The pipeline starts by itself every Monday at 6:00 with nobody involved:

If there are not enough confirmed facts for the topic, the article is not written and the topic is dropped.

The client left a 5/5 review on the independent Slovak freelance portal Jaspravim.sk