PromptMN: Pseudo Prompting Language

The paper introduces PromptMN, a pseudo-prompting domain-specific language designed to enhance the clarity and structure of prompts used with generative AI models. By utilizing compact, %-prefixed typed directives to annotate roles, goals, and constraints, PromptMN allows for semantic resolution and flexible prompt construction, improving the interpretability of user intentions. Initial evaluations with models like Claude Fable 5, Claude Opus 4.8, Gemini 3.1 Pro, and GPT-5.5 demonstrate successful resolution of complex instructions without fine-tuning, indicating its potential to streamline human-AI interactions and reduce ambiguity in software development workflows.