API Reference
Organized by module family with representative endpoints, input expectations, equations, constants, and integration notes.
deepsafety.tech
Deep Safety provides a verifiable computational layer for consequence analysis, hazard evaluation, and operational decision support. One consistent model contract spans notebooks, APIs, browser applications, GIS workflows, and agent systems.
Overview
Material properties and operating conditions are treated as upstream dependencies. Source models generate release rate and phase state; dispersion and fire/explosion models convert those outputs into consequence metrics; hazard-evaluation workflows then consume the same outputs for HAZOP, FMEA, and What-If studies. This makes the chain auditable and consistent across tools.
Modules
Organized by module family with representative endpoints, input expectations, equations, constants, and integration notes.
Step-by-step examples for package users, API integrators, and browser-local users who want fast onboarding.
Concrete patterns for pipeline leaks, toxic releases, siting screens, sign-triggered screening, and hazard-study support.
Use the README page for installation, packaging, Docker, MCP, and project-level details outside the API reference itself.
Highlights
Materials, source terms, dispersion, fire, explosion, effect, prevention, relief, and hazard-evaluation modules are already exposed through the same API surface.
Equations, constants, and physical meaning are explicit in responses and docs. Constants can be overridden per request for sensitivity studies and what-if analysis.
Use the same model contract in browser-local mode, Python notebooks, FastAPI services, Docker deployments, and MCP tools without rewriting calculation logic.