Create a single LaTeX worksheet for Singapore secondary school math with concise notes, aligned new questions based on a named reference exercise, and very concise end-of-document solutions.
Detect and remove known APEX targets from born-digital PDFs — logos, cartoons, and Must Do/Challenging badges — whether rendered as image XObjects or vector paths. Use when users want these recurring elements removed while preserving non-target content.
Create and run Python scripts that submit OpenAI Responses API requests in background mode, poll response status periodically, and stream progress with unbuffered stdout so Codex can monitor long-running jobs in the harness. Use when tasks require background=True calls, response_id polling, or live progress monitoring from python -u output.
Convert PDF textbook pages into LaTeX by rendering pages to PNG at 300 DPI, uploading exactly one full page image per OpenAI Responses API request with background=True, and polling by response id until terminal status. Use when users ask for OCR/transcription of math-heavy pages and want API-side background work with strict one-page-per-request processing.
Orchestrate simultaneous in-flight background-mode LaTeX transcription jobs by rendering a PDF to single-page PNGs at 300 DPI or consuming prepared single-page images, submitting many one-image Responses API requests with background=True, persisting response ids in a manifest, polling them to completion, extracting LaTeX snippets in order, and resuming after interruption. Use when users want real parallel background execution inside Codex rather than a single-image runner.
Convert PDF textbook pages into LaTeX by rendering pages to single-page PNGs at 300 DPI and uploading one image per synchronous OpenAI Responses API request. Use when users ask for OCR/transcription of math-heavy pages, full prose-plus-math LaTeX extraction, or reliable one-page-at-a-time processing with ordered snippet assembly and pdflatex compilation.
Detect and remove watermarks, headers, and footers from born-digital PDFs by targeting tagged /Artifact BDC/EMC blocks in the content stream. Use when users ask to remove watermarks, strip headers/footers, clean up PDF page furniture, or identify repeating overlay elements across PDF pages. Requires pikepdf and pymupdf (fitz).