danyuchn/asd-ste100-skill

Project overview

It repurposes the aerospace STE controlled-language standard—originally built so maintenance instructions cannot be misread—for an AI-agent reader with no human in the loop, which addresses a specific failure mode of agent-to-agent communication.

Project type

AI Agent · Prompt Engineering

Use cases

Documents & Office

Deployment

Refer to project documentation

License

MIT

Best for

• Developers and AI engineers who produce agent outputs, tool descriptions, error messages, or inter-agent instructions read by downstream AI agents with no human available to clarify ambiguity.
• Teams that want text rewrites that keep every fact, condition, and scope qualifier from the original rather than summarizing it away.

Key capabilities

• Rewrites dense, ambiguous English into ASD-STE100 Simplified Technical English.
• Flags every rule violation sentence-by-sentence, including ambiguous word choice, present-perfect/complex tense, passive voice with an unclear actor, multi-instruction sentences, oversized noun clusters, dropped words, sentences over length, phrasal verbs, nominalized actions, semicolons, hedge stacks, and marketing adjectives.
• Strict mode covers procedures, error messages, and tool descriptions; STE-flavored mode covers READMEs, PR descriptions, and explanatory prose.
• Rewrites each flagged sentence without dropping any fact, condition, or scope qualifier from the original.
• Outputs only the rewritten text—no preamble, mode announcement, or change summary—with a one-line Kept as-is note when something was deliberately left unsimplified.
• When asked to show the diff or which rules were broken, outputs a before/after table naming each rule instead.
• A deterministic Python linter that checks semicolons, a short list of soft phrasal verbs, nominalizations, marketing adjectives, passive voice, present-perfect forms, long sentences, synonym rotation, and dangling conjunctions in supported list items.
• Includes examples/before-after.md with more before/after illustrations and a linter-edge-cases.md test fixture.

Limitations and risks

• Does not reproduce ASD's official ~900-word approved dictionary; it applies the underlying principle rather than a fixed word list.
• The linter checks structural patterns only and does not verify that a rewrite preserved meaning.
• ASD-STE100 Issue 9 permits reproduction only with ASD's written authority or by eight listed organisation categories this project does not belong to, so compliance conclusions about the underlying standard should be verified against that restriction.

Getting started

• Run npx skills add danyuchn/asd-ste100-skill from the project root (or a single git clone), then paste text and ask Claude to apply STE100 to it.

Evidence and sources

• README: A Claude Code skill that rewrites dense, ambiguous English into [ASD-STE100 Simplified Technical English](https://www.asd-ste100.org/) (STE)
• README: an **AI agent** parsing another agent's output, a tool description, an error message, or an inter-agent instruction, with no human in the loop to resolve ambiguity.
• README: Rewrites each flagged sentence — without dropping any fact, condition, or scope qualifier from the original.
• README: Picks a mode. **Strict** covers procedures, error messages, and tool descriptions. **STE-flavored** covers READMEs, PR descriptions, and explanatory prose.
• README: Flags every rule violation sentence-by-sentence: ambiguous word choice, present-perfect/complex tense, passive voice with an unclear actor, multi-instruction sentences, oversized…

Source updated at: 2026-10-10T18:11:24Z

Open original GitHub project