PolyGnosis 2.0: Enhancing LLM Reasoning via Agentic Harness Engineering for Polymarket and OSINT Insight Extraction

2026-05-25

AI Summary

This study addresses the challenge of extracting predictive signals in high-noise financial environments by leveraging the “perspective mismatch” between Polymarket sentiment and global media narratives from GDELT as a high-alpha trading signal. The authors propose a multi-agent system that integrates anomaly detection, open-source intelligence, and tailored coordination mechanisms to systematically evaluate the efficacy of “steering engineering” techniques—including reflective loops, tool use, chain-of-thought reasoning, and divide-and-conquer strategies—in financial forecasting. Their findings reveal that structured divide-and-conquer is critical for performance, while unconstrained reflection often induces logical drift and exposes pervasive consensus bias. The work identifies a Pareto-optimal configuration that maintains expert-level analytical accuracy while substantially reducing latency and token consumption, outperforming human benchmarks.

Abstract

This paper introduces PolyGnosis 2.0, a pioneering multi-agent architecture designed to extract predictive intelligence by synthesizing Polymarket anomaly signals with global Open Source Intelligence (OSINT) streams, specifically Global Database of Events, Language, and Tone (GDELT). We define and target "Perspective Mismatches", the narrative divergence between Polymarket sentiment and global media flows, as high-alpha trading signals. Moving beyond generic agentic superiority, we rigorously quantify the efficacy of "Harness Engineering" techniques, including reflection loops, tool-calling, divide-and-conquer partitioning (D&C), and chain-of-thought (CoT), within high-noise financial domains. Our empirical evaluation against human-expert benchmarks reveals that while structural partitioning is mandatory for multi-dimensional alignment, unconstrained terminal reflection actively induces logical drift. Furthermore, we identify a pervasive "consensus bias" across all agent configurations during narrative reasoning, necessitating deterministic validation. Ultimately, we isolate a Pareto-optimal configuration that achieves professional-grade analytical precision while minimizing latency and token overhead, providing a robust blueprint for autonomous intelligence in prediction markets.

Problem

• Perspective Mismatches
• Polymarket
• OSINT
• prediction markets
• narrative reasoning

Innovation

• Harness Engineering
• Perspective Mismatches
• Multi-agent Architecture
• OSINT Integration
• Consensus Bias