IAbuElRuzz ICD PDF Analysis
Use this skill for ICD/PDF/documentation work.
Source Handling
- Identify each source document by filename, title, issue/revision, date and role before comparing.
- Extract text with
pdftotext -layoutwhen building tables or summaries. - Keep generated deliverables in the user-requested folder.
- Keep source Markdown next to generated PDFs when creating deliverables.
- In this repository, default ICD source PDFs are in
ICD_Documents/. - In this repository, default G1/G2 difference, similarity and status outputs are in
ICD_Documents/g1_g2_differences/.
Analysis Rules
- Separate confirmed facts from inferences.
- Mark classified, TBD, deleted, spare, reserved, dummy, or not-public items explicitly.
- For page/frame tables, include channel/interface, direction, page or frame size, symbol/bit size, frame/sub-frame timing, interval, fields and source document.
- Call out classified-annex or external-reference limits when public details are not available.
- Do not infer a missing layout or protocol detail unless clearly stated as an inference.
Output Format
- Prefer Markdown source plus searchable PDF.
- Keep source
.mdnext to the generated.pdf. - Use landscape PDF layout for wide comparison tables.
- Include an executive summary, detailed tables, and a missing/open-items register for status documents.
Useful Commands
- List local ICD PDFs:
find ICD_Documents -maxdepth 2 -type f -name '*.pdf' | sort - Extract PDF text:
pdftotext -layout 'path/to/document.pdf' /tmp/document.txt - Extract a local ICD PDF:
pdftotext -layout 'ICD_Documents/[AD.xx].pdf' /tmp/adxx.txt - Inspect metadata:
pdfinfo 'path/to/document.pdf' - Search extracted text:
rg -n "frame|subframe|sub-frame|page|alert|uplink|downlink|band|TBD|classified|reserved" /tmp/document.txt