Itu T Recommendation Reviewer
Model-agnostic AI agent skill (open Agent Skills format) that reviews draft ITU-T Recommendations against the ITU-T Author's Guide, presentation rules and quality checklist. Works with any LLM.
openitu · AI review on file
Installation
Read the source and dependencies before running a command. Copying does not install or execute anything.
npx skills add openitu/itu-t-recommendation-reviewerAbout this skill
ITU-T Recommendation Reviewer
A model-agnostic AI agent skill that reviews draft ITU-T Recommendations (including ITU-T | ISO/IEC common texts, amendments, and corrigenda) for compliance with the official ITU-T drafting rules — before the text goes to TSB or is proposed for consent, determination, or approval.
The skill is written entirely in plain Markdown, so any large language model or AI agent can use it — load the files into the model's context, or install them in any agent framework that supports the open Agent Skills format (SKILL.md + reference files).
It encodes the drafting and presentation rules from three authoritative sources:
- ITU-T Editing Guidelines — Author's guide for drafting ITU-T Recommendations (06/2023) — the primary rulebook for pure ITU-T texts (ITU permalink)
- Rules for presentation of ITU-T | ISO/IEC common text (09/2014) — overrides the Author's guide for joint ITU-T | ISO/IEC documents
- Recommendation ITU-T A.1 (09/2019) — ITU-T working methods (ITU permalink)
The ITU source documents themselves are not redistributed in this repository. They are © ITU and available free of charge from the links above and the ITU-T website.
What it does
Given a draft Recommendation (as .docx, PDF, or pasted text), the skill instructs the model to:
- Classify the document type first — pure ITU-T text, ITU-T | ISO/IEC common text, or amendment/corrigendum — and apply the correct rulebook.
- Run a six-pass review: skeleton & mandatory elements → boilerplate wording → references & citations → definitions & abbreviations → figures/tables/equations/notes → quality checklist & language.
- Produce a structured review report with an overall verdict (READY / MINOR REVISION / MAJOR REVISION), findings graded Critical / Major / Minor, each pinpointed to a clause and c