InfoSys Cambridge Hackathon: Multi-Agent Fact-Checking for Social Media
- agent-framework
MCP Hackathon Repository
About
What is InfoSys Cambridge Hackathon: Multi-Agent Fact-Checking for Social Media?
FactCheckBot is a multi-agent system built using the Model Context Protocol (MCP) that identifies, verifies, and responds to factual claims on social media platforms like Twitter (X). It decomposes fact-checking into modular agents responsible for specific reasoning steps, designed to combat misinformation with speed, context, and transparency.
How to use InfoSys Cambridge Hackathon: Multi-Agent Fact-Checking for Social Media?
The system takes a tweet or thread as input and runs a workflow of six agents sequentially: ClaimExtractorAgent, RetrieverAgent, VerifierAgent, CrowdSignalAgent, ExplainerAgent, and PosterAgent. The README does not specify installation or configuration steps beyond the agent flow.
Key features of InfoSys Cambridge Hackathon: Multi-Agent Fact-Checking for Social Media
- Six modular agents, each handling a distinct fact-checking task
- Uses Model Context Protocol for shared memory and agent coordination
- Falls back on crowdsourced consensus for breaking or unverified events
- Generates explainable verdicts with sources and reasoning
- Supports automatic reply posting to original tweets
Use cases of InfoSys Cambridge Hackathon: Multi-Agent Fact-Checking for Social Media
- Verifying breaking news claims before major outlets report them
- Detecting and correcting misinformation about current events or science
- Providing fallback credibility scores when traditional sources are unavailable
- Publishing structured, human-readable replies automatically to original tweets
FAQ from InfoSys Cambridge Hackathon: Multi-Agent Fact-Checking for Social Media
What platforms does FactCheckBot support?
The system is designed for social media platforms like Twitter (X). It extracts claims from tweets or threads and posts replies.
How does the system handle claims that can't be verified through traditional sources?
It uses CrowdSignalAgent to search for geolocated, independent social media posts within a time window, evaluate consensus, and provide a fallback credibility score.
What sources does the RetrieverAgent search?
It searches verified sources such as Wikipedia, arXiv, news APIs, and others mentioned in the README.
How does the system produce explainable verdicts?
The ExplainerAgent summarizes the verdict in human-readable language with reasoning, and the PosterAgent includes sources and explanations in the reply.
What is the role of the Model Context Protocol (MCP)?
MCP defines a shared memory structure that allows dynamic coordination and context-aware reasoning across all agents.
Details
- Author
- denizakansoy
- GitHub stars
- 1
- Category
- agent-framework
- Repository
- denizakansoy/mcphack
InfoSys Cambridge Hackathon: Multi-Agent Fact-Checking for Social Media
FactCheckBot is a multi-agent system designed to identify, verify, and respond to factual claims on social media platforms like Twitter (X). Built using the Model Context Protocol (MCP), the system decomposes the complex task of fact-checking into modular, intelligent agents — each responsible for a specific decision or reasoning step.
---
What It Does
Given a tweet or thread, FactCheckBot:
1. Extracts factual claims
2. Searches reliable external sources
3. Evaluates the claim based on supporting or contradicting evidence
4. Falls back on crowdsourced consensus for recent events
5. Generates a concise, explainable verdict
6. Posts a correction or clarification as a reply
This system is designed to combat misinformation with speed, context, and transparency.
---
Agent Roles
| Agent Name | Description |
|----------------------|-------------|
| ClaimExtractorAgent | Detects and isolates factual claims from tweets |
| RetrieverAgent | Searches verified sources (Wikipedia, arXiv, news APIs, etc.) |
| VerifierAgent | Compares the claim to retrieved evidence and makes an initial judgment |
| CrowdSignalAgent | In breaking news cases, checks for regional consensus across social media |
| ExplainerAgent | Summarizes the verdict in human-readable language, with reasoning |
| PosterAgent | Posts a reply to the original tweet, including sources and explanations |
All agents operate on a shared memory structure defined by the Model Context Protocol, allowing dynamic coordination and context-aware reasoning.
---
Example Workflow
Input Tweet:
> “Massive explosion reported in Ankara right now.”Agent Flow:
1.ClaimExtractorAgent detects a location-based factual claim.
2. RetrieverAgent finds no results in verified sources — too recent.
3. CrowdSignalAgent clusters 10+ independent geolocated users in Ankara posting similar claims.
4. VerifierAgent assigns a tentative credibility score with fallback reasoning.
5. ExplainerAgent formats the reply:
> “🚨 Developing: This event is not yet reported by major outlets, but multiple users in Ankara are sharing photos and consistent details. Treat as likely true, pending confirmation. [Context]”
6. PosterAgent publishes the reply.
---
Real-Time Consensus via CrowdSignalAgent
In cases where a claim cannot be confirmed through traditional sources (e.g., breaking news or remote events), the system uses CrowdSignalAgent to:
- Search for tweets matching the claim
- Filter for geolocated, domain-relevant, or independent sources
- Evaluate consensus within a time window
- Provide a fallback credibility score if traditional verification is unavailable
This helps prevent false negatives and enhances trustworthiness in time-sensitive cases.
---
