CheckMate - quick and easy-to-use fact-checking assistant that combats misinformation

This is a submission for the Agent.ai Challenge: Full-Stack Agent (See Details)

What I Built

CheckMate is a fact-checking assistant built on Agent.ai. It helps users quickly verify suspicious headlines, viral claims, or rumors by querying reliable fact-check sources (e.g., Google Fact Check Tools). With a simple chatbot interface, CheckMate encourages critical thinking and combats misinformation. It’s envisioned as a handy tool for anyone who wants to quickly confirm or debunk circulating claims, especially in fast-paced social media environments.

Demo

You can test the CheckMate agent at:
https://agent.ai/agent/checkmate

Here’s how it works in action:

1. User inputs a suspicious claim or headline.
2. CheckMate queries the Google Fact Check API via a serverless function.
3. The Agent returns summarized fact-check results or indicates no matches.

Agent.ai Experience

Building CheckMate with Agent.ai was an exciting journey. The Flow Builder made it straightforward to create the conversation structure. I utilized the Call Serverless Function action to connect with AWS Lambda function (written in Python) that queries the Google Fact Check Tools API.

Delightful Points

The visual Flow Builder helped us quickly link user input to our serverless logic.
Testing the agent in real-time allowed for rapid iteration.

Challenges

Handling the prefix in Google’s JSON response ()]}'\n) required custom parsing.
Ensuring the Lambda environment had the right libraries or using only standard Python libraries was a bit tricky at first.

In the end, CheckMate can streamline fact-checking into a simple chatbot flow, demonstrating how AI Agents can help combat misinformation and empower users to confirm what they see online.