skills.rest
xyzxyq
xyzxyqCOMMUNITY
Explore all skills by this author →
GitHub

docx-course-report-writer

Generate and verify Chinese academic DOCX reports with evidence tracking, Word field updates, and PDF page-level QA.

52|4|Updated Jun 8, 2026
Search Tags:#docx#evidence tracking#word automation#chinese academic writing#course report#gb/t 7714#pdf qa
One-click install
npx skills add https://github.com/xyzxyq/docx-course-report-writer --skill docx-course-report-writer-xyzxyq
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: docx-course-report-writer
Source: https://github.com/xyzxyq/docx-course-report-writer
Command: npx skills add https://github.com/xyzxyq/docx-course-report-writer --skill docx-course-report-writer-xyzxyq

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires python-docx, Pillow, and includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve? Chinese coursework reports repeatedly fail at delivery time: stale template residue, unupdated Word tables of contents, unverifiable references, unreadable diagrams, and fabricated experiment evidence. This Skill turns report writing into a planned, evidence-first, reviewed, and verifiable DOCX/PDF workflow instead of a one-shot text generation. ## Core Features & Use Cases - Evidence-first report generation: Builds a fact ledger from real code runs, logs, screenshots, and data before drafting conclusions, and locks a verified GB/T 7714-2015 reference metadata ledger before writing citations. - Word-native document engineering: Uses real Word heading styles, automatic TOC fields, Word COM field updates and PDF export on Windows, template-fidelity preservation for user-provided DOCX templates, and page-level PDF rendering QA with blank-page detection. - Actor/Critic review loop: Runs at least two mandatory Actor/Critic cycles that audit the actual DOCX/PDF artifacts, plus diagram Review And Revise gates for TikZ figures, flowcharts, screenshots, and controlled AI-generated concept images. - Use Case: Given an assignment brief, source code, and a course template, produce a Chinese lab report with real terminal screenshots, formal figure captions, superscript numeric citations, an updated TOC, and a rendered-PDF-verified final deliverable. ## Quick Start Ask Codex to use the docx-course-report-writer skill to generate a Chinese lab report from your assignment file, source code, and template, requiring real run evidence, screenshots, and both DOCX and PDF output.

Frequently Asked Questions about docx-course-report-writer

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate a Chinese course report DOCX with Codex?▼

Install the skill into your Codex skills directory and invoke it with your assignment brief, source code, and optional DOCX template. It builds an evidence plan, drafts the report, assembles the DOCX with real heading styles and an automatic TOC, then exports and visually verifies a PDF.

How does the skill update Word TOC and page numbers?▼

It uses real Word heading styles and automatic TOC field codes rather than static text. On Windows it prefers Word COM automation via the update_word_fields.ps1 script to update fields, refresh the TOC, and export the final PDF.

Can I use my own DOCX template for the report?▼

Yes. A user-provided DOCX template takes priority and is treated as the authoritative layout source: the skill copies it first, replaces placeholders and sample content in place, and preserves its cover, styles, headers, and page setup. Without one, it uses the integrated default template.

Does it support AI-generated images in reports?▼

Yes, but only after explicitly asking whether text-to-image is enabled and the maximum image count. AI images are limited to conceptual or explanatory use, must pass a prompt-card and density review, and can never replace real experiment screenshots or data plots.

What citation format does the report use?▼

References default to GB/T 7714-2015 sequential numeric style with superscript bracketed in-text citations. A reference metadata ledger is verified against DOI or authoritative sources before drafting, so unverified or hallucinated references are blocked.

Why does the workflow require two Actor/Critic review cycles?▼

Most report defects appear after the first draft: stale template content, broken TOC fields, weak evidence, or unreadable diagrams after scaling. Two mandatory cycles ensure the Critic audits the actual regenerated DOCX/PDF artifact, catching regressions from field updates and late edits.

Related Skills

Explore More Education & Research Skills→
recallnetrecallnet
1

grade-experiment

Grade experiment documents by applying a rubric and updating frontmatter and narrative notes.

Official
Advanced
MIKUZ12MIKUZ12
20

exampass

Convert lecture PPTX, DOCX, PDF, and screenshots into chapter guides and interactive HTML tests.

Community
Advanced
NikhilVijayakumarNikhilVijayakumar

paper-reviewer

Cross-check academic drafts against YAML plans and source data.

Community
Advanced
Pixel-Process-UGPixel-Process-UG
1

docx-processing

Generate and manipulate Word documents with python-docx and docxtpl.

Official
Advanced