Higher Categorical Semantics of an Embedding-Financed Proxy Company
Version 1.0 - 25 October 2026 - Anonymous author

CONTENTS
  manuscript.md          Authoritative manuscript, Markdown with literal LaTeX.
  manuscript.tex         Generated standalone LaTeX for the PDF.
  embedding-proxy-company.pdf   Typeset manuscript.
  verify_examples.py     Exact finite checks, Python standard library only.
  verification.json      Recorded verification output.
  submission.json       Exact anonymous submission fields; no credentials.
  header.tex            PDF typography configuration.
  SHA256SUMS             Checksums of the files above and this README.

CHECK THE EXAMPLES
  python3 verify_examples.py --output verification.json

BUILD THE PDF
  Requires Pandoc, XeLaTeX, Latin Modern Roman and Latin Modern Math.
  pandoc manuscript.md --standalone --pdf-engine=xelatex \
    -V mainfont="Latin Modern Roman" -V mathfont="Latin Modern Math" \
    -V fontsize=11pt -V geometry:margin=0.85in \
    -H header.tex -o embedding-proxy-company.pdf

SCOPE
  This is a mathematical specification, with finite synthetic checks. It is
  not a proxy service implementation, proof-assistant formalization, privacy
  certification, legal conclusion, benchmark, or revenue forecast. General
  results have written mathematical proofs. The assumed buyer payments in
  the existence construction demonstrate consistency rather than demand.

  "Embedding retention" includes declared metadata in the information
  boundary. It does not imply irreversibility, anonymity, or safety of selling
  exact vectors. The private-release interface has a separate guarantee.

PUBLICATION
  Prepared for submission under the byline Anonymous author. The human is
  responsible for direction and authorization. OpenAI ChatGPT assisted with
  formulation, retrieval, writing, and checks. The submission receipt is
  separate from this public artifact package. Editorial review and approval
  are distinct from successful submission.
