This tool provides neuromorphic threat intelligence for AI agents, functioning as an MCP server that applies 8 brain-inspired and mathematical frameworks to cyber threat detection. It pulls live data from 15 sources including vulnerability databases, internet scanning infrastructure, sanctions registries, social signals, and code repositories. It enables AI assistants like Claude, Cursor, and any MCP-compatible client to reason about threats using spiking dynamics, temporal learning, and topological structure. Each tool call fires parallel requests across multiple intelligence feeds, running rigorous algorithms like leaky integrate-and-fire spiking networks or hypergraph attack grammars, and returns structured JSON with an interpretable conclusion, going beyond simple API wrappers to provide multi-source reasoning with quantitative and clear threat assessments.
Key Features
01Hypergraph attack grammar for infrastructure attack path analysis
02Severity normalisation across diverse threat data formats (CVSS, strings, raw scores)
03Plain-language interpretations accompanying all numerical outputs
04Leaky integrate-and-fire (LIF) spiking network for real-time anomaly detection
05STDP learning with Tracy-Widom edge for temporal threat campaign attribution
060 GitHub stars
