Prediction Market Analyst Team
Create a specialized team of AI agents using CrewAI to analyze real-time data from prediction markets (e.g., sports betting, event futures like Kalshi). The agents, powered by Claude Opus 4.1, will collaborate to identify market trends, perform sentiment analysis on related news, predict potential market movements, and generate concise, actionable market commentary or alerts. The challenge emphasizes implementing robust agent-to-agent (A2A) communication using a custom A2A protocol for secure and structured data exchange between agents. Agents will also leverage MCP (Multi-Agent Coordination Protocol)-enabled tools to access external data sources like market APIs and news aggregators. The system should showcase hybrid reasoning, combining instant analysis for quick alerts with deeper, extended reasoning for comprehensive reports.
What you are building
The core problem, expected build, and operating context for this challenge.
Orchestrate a CrewAI team of agents with Claude Opus 4.1, using a custom A2A Protocol and MCP tools for real-time prediction market analysis and commentary.
How work is evaluated
The evaluation will assess the CrewAI team's ability to accurately analyze prediction market data, identify trends, generate relevant commentary, and effectively use A2A communication and MCP tools.
Challenge datasets
Public datasets, plus the private datasets you have access to.
What you should walk away with
Develop hybrid instant/deep reasoning capabilities with Claude Opus 4.1 for varying analysis depths.
Design and integrate MCP-enabled tools for real-time data access (market APIs, news feeds).
Implement a custom A2A Protocol for secure, structured agent-to-agent communication.
Orchestrate role-based agent teams using CrewAI for complex collaborative tasks.
Reference links and supporting material
Sample data for 2 tasks
Challenge setup
The evaluation will assess the CrewAI team's ability to accurately analyze prediction market data, identify trends, generate relevant commentary, and effectively use A2A communication and MCP tools.
Challenge input
{"market_data_snapshot": "json_string", "news_articles": "list_of_strings", "timeframe": "string (e.g., '24 hours')"}
Agent execution
No agent tool is listed for this challenge.
Evaluated output
{"identified_trend": "string", "predicted_movement": "string (e.g., 'up', 'down', 'stable')", "confidence_score": "float", "rationale": "string"}
- Ensure inter-agent communication adheres to the defined A2A protocol structure and security.
- Verify agents successfully interact with external APIs via MCP-enabled tools.
- Ensure generated commentary and analysis are directly relevant to the input data and market context.
- Analysis Factual Accuracy target: 0.85
- Python execution harness
Summary shows the first task, up to three tools, and up to three evidence items. Open the technical recipe to view all configured items.
View technical recipe
Configured tools
No tool records are attached.
Evaluation contract
- Ensure inter-agent communication adheres to the defined A2A protocol structure and security.
- Verify agents successfully interact with external APIs via MCP-enabled tools.
- Ensure generated commentary and analysis are directly relevant to the input data and market context.
Metrics
- Analysis Factual Accuracy · Target 0.85
- Commentary Coherence and Professionalism · Target 4.5
- Agent Collaboration Efficiency · Target 5
Workflow tasks
- Task 1
Input: {"market_data_snapshot": "json_string", "news_articles": "list_of_strings", "timeframe": "string (e.g., '24 hours')"}
Output: {"identified_trend": "string", "predicted_movement": "string (e.g., 'up', 'down', 'stable')", "confidence_score": "float", "rationale": "string"}
Execution harness: Python
- Task 2
Input: {"event_description": "string", "market_impact_analysis": "string"}
Output: {"market_commentary": "string", "actionable_insights": "list of strings"}
Execution harness: Python
Recipe freeze marker
No recipe freeze time is recorded.