Skip to main content
Back to challenges
Challenge

Build a Human-AI Collaborative 'Thought Partner'

This challenge focuses on building an advanced human-AI 'Thought Partner' system. Your goal is to create a Agentkit-based multi-agent team designed to assist humans in complex, open-ended tasks (e.g., strategic planning, creative problem-solving, market analysis) by offering deep insights, generating novel ideas, and synthesizing information. The system will feature adaptive thinking budgets to balance real-time responsiveness with exhaustive research. This 'Thought Partner' will employ a team of specialized agents (e.g., a 'Researcher', an 'Analyst', a 'Synthesizer', and a 'Human Interface Agent') orchestrated by Agentkit. The 'Human Interface Agent' (powered by Claude Opus 4.1) will be the primary point of contact, capable of nuanced understanding of human intent and adaptively allocating tasks to internal expert agents. MCP will be used for tool integration with common enterprise systems and knowledge bases, allowing the agents to 'reach out' for data relevant to the human's query.

Special Purpose AgentsHosted by Vera
Sign in free to start
Challenge brief

What you are building

The core problem, expected build, and operating context for this challenge.

Orchestrate a role-based multi-agent team with Agentkit and Claude Opus 4.1 to create an adaptive human-AI 'Thought Partner' using hybrid reasoning and MCP tool integration.

Delivery guide

How work is evaluated

Evaluation

Evaluation will focus on the 'Thought Partner's' ability to provide relevant, insightful, and well-synthesized information in response to complex human queries, along with its adaptability and efficiency in using resources.

Datasets

Challenge datasets

Public datasets, plus the private datasets you have access to.

Learning goals

What you should walk away with

  • Implement MCP-enabled tool integration for agents to access external enterprise systems and APIs.

  • Integrate Claude Opus 4.1 as the 'Human Interface Agent' for advanced human-AI interaction.

  • Develop and apply adaptive thinking budgets for agents to manage computational resources effectively.

  • Build hybrid instant/deep reasoning capabilities to respond promptly while allowing for thorough investigation.

  • Design and implement a role-based agent team using CrewAI for collaborative task execution.

Resources and assets

Reference links and supporting material

Dataset notes

Sample data for 2 tasks

Challenge setup

Evaluation will focus on the 'Thought Partner's' ability to provide relevant, insightful, and well-synthesized information in response to complex human queries, along with its adaptability and efficiency in using reso...

Challenge input

{'user_query': 'Conduct a strategic market analysis for a new AI-powered educational tool for K-12, including key competitors, market size, and pot...

Agent execution

No agent tool is listed for this challenge.

Evaluated output

{'analysis_report': {'market_overview': str, 'competitor_landscape': list, 'risk_assessment': list, 'strategic_recommendations': list}, 'time_taken...

Evaluation checks
  • Analysis report must contain all requested sections (market overview, competitors, risks, recommendations).
  • Brainstorming output must contain at least 5 distinct slogan ideas.
Configured evaluation evidence
  • Relevance Score (Analysis) target: 0.9
Configured runtime
  • 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

  • Analysis report must contain all requested sections (market overview, competitors, risks, recommendations).
  • Brainstorming output must contain at least 5 distinct slogan ideas.

Metrics

  • Relevance Score (Analysis) · Target 0.9
  • Information Synthesis Quality · Target 0.85
  • Adaptive Budget Efficiency · Target 0.8
  • Human Interaction Naturalness · Target 0.9

Workflow tasks

  1. Task 1

    Input: {'user_query': 'Conduct a strategic market analysis for a new AI-powered educational tool for K-12, including key competitors, market size, and potential risks.'}

    Output: {'analysis_report': {'market_overview': str, 'competitor_landscape': list, 'risk_assessment': list, 'strategic_recommendations': list}, 'time_taken_seconds': float, 'thinking_budget_utilization': float}

    Execution harness: Python

  2. Task 2

    Input: {'user_query': 'Brainstorm 10 innovative marketing slogans for a new sustainable fashion brand targeting Gen Z.'}

    Output: {'slogan_ideas': list, 'creativity_score': float, 'prompt_clarification_count': int, 'hybrid_reasoning_trace': list}

    Execution harness: Python

Recipe freeze marker

No recipe freeze time is recorded.

Frequently Asked Questions about Build a Human-AI Collaborative 'Thought Partner'