Overview

This Claude Code skill transforms complex, ambiguous English into ASD-STE100 Simplified Technical English (STE). Originally developed for the aerospace and defense industry to ensure critical instructions are never misread, this skill repurposes STE's rigorous rules for AI agent communication. It tackles the challenge of AI agents needing to parse information—like tool descriptions, error messages, or inter-agent instructions—without human clarification.

Key Features

• Dual Modes: Operates in Strict mode for procedures and error messages, and STE-flavored for less formal prose like READMEs, maintaining sentence discipline without strict vocabulary control.
• Comprehensive Rule Enforcement: Identifies and rewrites sentences violating STE principles, including ambiguous wording, complex tenses, passive voice, multi-instruction sentences, long sentences, noun clusters, and more.
• Meaning Preservation: Rewrites text meticulously, ensuring no facts, conditions, or scope qualifiers are lost. If precision requires a longer phrasing, it's used and flagged.
• Actionable Feedback: Optionally provides detailed diffs, highlighting specific STE rule violations per sentence.
• No Preamble Output: Delivers rewritten text directly, with an option to explicitly note unsimplified segments (Kept as-is:).
• Flexible Installation: Available via npx skills add for quick project integration or by cloning the repository for deeper use.

Typical Use Cases

This skill is ideal for scenarios where clarity and lack of ambiguity are paramount for machine interpretation:

• Agent-to-Agent Communication: Ensuring instructions and data shared between AI agents are perfectly understood.
• Tool and Function Descriptions: Creating unambiguous documentation that downstream agents can reliably parse.
• Error Message Generation: Producing clear, actionable error messages that AI systems can process without misinterpretation.
• System Prompts: Crafting precise instructions for AI agents to follow.
• Inter-Agent Instructions: Defining clear directives for complex multi-agent workflows.