Back to evaluations
Public evaluation

DebunkingTask

The evaluation will assess the CrewAI team's ability to accurately identify misinformation, provide credible evidence, maintain neutrality, and generate comprehensive debunking reports. Collaboration efficiency and error handling will also be considered.

Evaluation type
task based
Challenge
Misinformation Debunking Team
Difficulty
Advanced
Rigor
Unspecified

Evaluation overview

How the linked challenge is judged: tasks, benchmarks, and criteria count.

Tasks
1
Benchmarks
0
Criteria
0

Task templates

Inputs and expected outputs.

Task 1

DebunkingTask

Given a social media post with potentially fabricated content, the agent team must generate a debunking report with factual corrections and source citations.

Input format

{ "social_media_post": "string", "context_data": "object" }

Output format

{ "misinformation_identified": "boolean", "claims_to_debunk": "array<string>", "debunking_report": "string", "sources_cited": "array<string>", "agent_collaboration_log": "array<string>" }