GitHub - 0xpili/simplified-technical-english: An Agent Skill that makes an LLM write in ASD-STE100 Simplified Technical English. Rules, approved word list, and a check tool.
Short summary
This is a GitHub project that teaches Claude (or other LLMs) to write in Simplified Technical English (STE), a controlled language used by aerospace and defense industries for clear, translatable technical documentation. It includes writing rules, an approved word list, and a Python checker tool to validate compliance.
- The skill enforces 53 writing rules and a dictionary of 869 approved words based on the official ASD-STE100 specification, making technical text shorter and easier to translate
- It includes a Python validation tool that catches common violations like overly long sentences, unapproved verbs, passive voice, and words outside the approved list
- The tool has real limitations—it can't verify if words are used with their approved meanings, and a human expert still needs to review critical documents for full specification compliance
Full summary
This is a GitHub repository that implements Simplified Technical English (STE) as a skill for Claude, an AI assistant. STE is a controlled language used in aerospace and defense documentation that enforces strict rules to make technical writing clearer and easier to translate. The repository contains the rules (53 of them), an approved word list (869 words), and tools to help writers follow the standard. The example given shows how STE transforms verbose, passive phrasing—"Prior to commencing the installation, it should be ensured that all components have been thoroughly inspected"—into direct, active language: "Before you start the installation, examine all the components for damage."
The repository includes several key files: SKILL.md provides the core instructions for the LLM, references contain the writing rules with examples and word substitutions, and a Python script (ste_check.py) validates text against STE requirements. The checker catches common violations like sentences exceeding word limits, paragraphs longer than six sentences, passive voice, unapproved verb forms, and words outside the approved dictionary. Users can integrate this into Claude by cloning the repository into their skills folder or manually adding the SKILL.md text to the system prompt.
The creator acknowledges important limitations. The checker can't catch all errors—it can't verify whether a word is being used in its approved sense, so human review remains necessary for critical documentation. The skill is based on ASD-STE100 Issue 7 from 2017, and since ASD releases updated specifications periodically, this implementation may not fully comply with newer versions. The word list and specification remain ASD's property, though the skill itself is released under MIT license.