Combat AI Slop: Multi-Agent System for Content Authenticity
The proliferation of AI-generated 'slop' on platforms like Reddit threatens to erode authenticity and create a self-perpetuating feedback loop where AI models train on AI-generated content. This challenge focuses on building a sophisticated multi-agent system designed to detect, analyze, and mitigate the impact of such content. Leveraging state-of-the-art 2025 LLMs and graph-based agent orchestration, you will develop a system capable of discerning subtle indicators of AI generation. The solution will integrate an MCP-enabled workflow to connect with hypothetical platform APIs for content flagging and moderation, ensuring that human-generated content remains the primary input for future AI training, thereby preserving the integrity of online communities.
What you are building
The core problem, expected build, and operating context for this challenge.
Build a LangGraph multi-agent system with GPT-5 and Claude Opus 4.1, using extended thinking and MCP tool integration to detect and mitigate AI-generated 'slop' on a content platform.
How work is evaluated
Evaluation will focus on the effectiveness of the multi-agent system in accurately identifying AI-generated content ('slop') and flagging it appropriately, as well as the robustness and scalability of the LangGraph workflow and MCP integration.
Challenge datasets
Public datasets, plus the private datasets you have access to.
What you should walk away with
Integrate advanced 2025 LLMs (GPT-5, Claude Opus 4.1) for nuanced text analysis.
Establish a feedback loop to refine detection models based on new data.
Employ extended thinking techniques with adaptive reasoning budgets for AI content detection.
Develop MCP-enabled agents for platform API interaction and moderation workflows.
Design and implement a graph-based multi-agent workflow using LangGraph.
Reference links and supporting material
Sample data for 2 tasks
Challenge setup
Evaluation will focus on the effectiveness of the multi-agent system in accurately identifying AI-generated content ('slop') and flagging it appropriately, as well as the robustness and scalability of the LangGraph wo...
Challenge input
{'texts': [{'id': '1', 'content': 'sample text 1'}, {'id': '2', 'content': 'sample text 2'}...]}
Agent execution
No agent tool is listed for this challenge.
Evaluated output
{'classifications': [{'id': '1', 'is_ai_slop': True, 'confidence': 0.95}, ...]}
- Verify that the LangGraph workflow executes without errors and follows the designed state transitions.
- Ensure that MCP-enabled tool calls are correctly formatted and reflect the moderation decisions.
- Slop Detection F1 Score 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
- Verify that the LangGraph workflow executes without errors and follows the designed state transitions.
- Ensure that MCP-enabled tool calls are correctly formatted and reflect the moderation decisions.
Metrics
- Slop Detection F1-Score · Target 0.85
- Reasoning Budget Efficiency · Target 1500
Workflow tasks
- Task 1
Input: {'texts': [{'id': '1', 'content': 'sample text 1'}, {'id': '2', 'content': 'sample text 2'}...]}
Output: {'classifications': [{'id': '1', 'is_ai_slop': True, 'confidence': 0.95}, ...]}
Execution harness: Python
- Task 2
Input: {'detection_results': [{'id': '1', 'is_ai_slop': True, 'action': 'flag'}, {'id': '2', 'is_ai_slop': False, 'action': 'none'}]}
Output: {'tool_calls': [{'agent': 'ModeratorAssistant', 'tool': 'flag_content', 'args': {'content_id': '1', 'reason': 'AI slop'}}, ...]}
Execution harness: Python
Recipe freeze marker
No recipe freeze time is recorded.