Show HN: ThreatStrata – A graph viewer for agent-generated threat models

ThreatStrata launched as a no-sign-up graph viewer that turns agent-generated threat models into interactive, encrypted links. Users copy a prompt into their own AI agent, point it at a repo, design doc, website, API spec, cloud config, system diagram, data schema or incident report, and the agent validates and submits a threat model that the viewer displays with threats, evidence and suggested checks. The tool targets threat modeling for AI agents by letting an existing agent map an architecture's threats rather than requiring a dedicated security product.

THREAT MODELLING FOR AI AGENTS Map the threats. Point your agent at an architecture. Get an interactive threat model. No sign-up required· Open the Threat Strata instructions and follow them. Build a threat model for: <describe system and architecture . Validate it, submit it, and return the viewer link. How it works Copy the agent prompt Use the prompt above to configure your agent. Point it at your system Point it at a repo, a design doc, a website, an API spec, a cloud config, a system diagram, a data schema, or an incident report. View the threat model Open the encrypted link and inspect threats, evidence, and suggested checks.