Prompt Lang
AGI Modeling Language — Storythought Architecture
Prompt Lang is a unified modeling language for AGI agents. It treats every Agent interaction as a programmable prompt, covering input/output modeling through the Universe Machine Interface (UMI), self-awareness via the Self Ontology Framework (SOF), declarative business logic with the RuleBased Engine (RBE), multi-agent pipelines with PPL/ippl protocol, and comprehensive benchmarking via ThoughtBench.
Core Architecture
- UMI — Universe Machine Interface for AGI input/output modeling
- SOF — Self Ontology Framework: physical, cognitive, resource, and skills ontology
- RBE — RuleBased Engine: declarative ECA-pattern rule-driven PNode execution
- PPL / ippl — Human-readable and model-readable prompt formats with pplc compiler
- Agent System — Multi-agent architecture with PPL communication, PNode nodes, RCP protocol
- Brain System — Composite synthesized multi-model with hierarchical planning
- ThoughtBench — IOE efficiency and ICS intelligence capacity benchmarks
- Security — Safe Prompts, RBAC, Master Prompts, prompt injection protection
GitHub · MachineSociety.com · MIT License
Enable JavaScript for the full interactive experience.