CredCheck

Inspiration

With misinformation spreading faster than memes, we were tired of seeing friends, family (and let's be honest — ourselves) getting duped by clickbait and shady news. We wanted a tool that could cut through the nonsense, check the facts, and throw a little shade while doing it. Thus, CredCheck was born — your sassy sidekick for truth on the web.

What it does

CredCheck is a Chrome extension that scans the news articles you’re reading and instantly checks their credibility. It analyzes the content for bias, tone, and factual reliability using AI, and cross-references it with trusted sources and reputation databases. It then gives you a clear, color-coded verdict — with a side of snark — so you know whether you’re reading solid journalism or hot garbage.

How we built it

We built a Chrome extension using HTML, JavaScript, and the Chrome Extensions API. When the user clicks the extension, it:

• Extracts the article content from the current page
• Sends the text to OpenAI's GPT-4o to analyze tone, political bias, and emotional language
• Cross-references the domain with public credibility APIs (like Media Bias/Fact Check and NewsGuard)
• Returns a summarized credibility score, bias rating, and an AI-generated “sassified” explanation We also used lightweight Firebase functions to store user reports and plan for community-based input.

Challenges we ran into

• Parsing article content cleanly from dynamic websites (curse you, infinite scroll and paywalls!)
• Making AI analysis feel fast enough for real-time use without exceeding token limits
• Integrating third-party trust data consistently (many APIs have inconsistent formats or limited access)
• Finding the right balance of sass and clarity — too funny, and people don’t take it seriously; too dry, and it loses its spark

Accomplishments that we're proud of

• We got real-time bias/tone analysis working directly in the browser
• The interface is clean, bold, and just snarky enough to be fun
• It’s surprisingly effective — we tested it on dozens of real articles, and it flagged the shady stuff accurately
• First-time building a full Chrome extension + integrating AI = ✅✅

What we learned

• How to build and publish a Chrome extension from scratch
• How to harness large language models (like GPT-4o) for natural language analysis and sentiment detection
• The importance of UX tone in trust-based tools — how you say it matters just as much as what you say
• That misinformation isn’t always obvious — nuance is hard, but AI + crowd input helps

What's next for CredCheck

• ✅ Add community voting: Let users rate articles and flag misinformation
• 🔍 Source-level breakdown: Show specific sentences flagged as biased or misleading
• 📱 Build a mobile-friendly version for on-the-go checks
• 📰 Partner with media literacy orgs to make CredCheck a real-world tool
• 💬 Integrate “fact-checked by CredCheck” share buttons to call out shady links in the group chat

We built CredCheck to make the truth loud, fun, and accessible — and we’re just getting started.

Built With

• chromeextensions-api
• figma
• firebase
• html/css
• javascript
• lottiefiles
• media-bias/fact-check-api
• openai:gpt-4o
• tidy-html-parser
• vs-code