一键导入
epub-pipeline-grc-zh-hans
// Use when creating or reviewing Ancient Greek to Simplified Chinese public-domain EPUB translation projects with this template.
// Use when creating or reviewing Ancient Greek to Simplified Chinese public-domain EPUB translation projects with this template.
Use this template when translating German public-domain or licensed source texts into Simplified Chinese EPUBs, especially literature, speculative fiction, nonfiction, social thought, and classic prose.
Use when creating or reviewing English to Simplified Chinese public-domain EPUB translation projects with this template.
Use this template when translating French public-domain or licensed source texts into Simplified Chinese EPUBs, especially literary nonfiction, natural history, travel, social thought, and classic prose.
Use when creating or reviewing Japanese to Simplified Chinese public-domain EPUB translation projects with this template.
Use when creating or reviewing Literary Chinese to modern Simplified Chinese public-domain EPUB translation projects with parallel source-modern passages and controlled annotations.
Use when translating public-domain or licensed books into Simplified Chinese, regardless of source language. Provides target-language quality rules for AI-assisted translation, including Chinese readability, style research, chapter gates, benchmark testing, literary revision, and EPUB-ready final text checks.
| name | epub-pipeline-grc-zh-Hans |
| description | Use when creating or reviewing Ancient Greek to Simplified Chinese public-domain EPUB translation projects with this template. |
Use this skill after books/scripts/create_book_project.py has copied template/epub_pipeline/common and overlaid template/epub_pipeline/grc-zh-Hans into books/zh-Hans/{number}_{book_id_slug}/.
使用本 skill 前,应先通过 books/scripts/create_book_project.py 把 template/epub_pipeline/common 复制到 books/zh-Hans/{number}_{book_id_slug}/,再覆盖复制 template/epub_pipeline/grc-zh-Hans。
If the work is classical science, mathematics, astronomy, or diagram/table-heavy, overlay template/epub_pipeline/profiles/classical-science-zh-Hans after this template.
如果作品属于古典科学、数学、天文学或图表/表格密集型作品,应在本模板之后叠加 template/epub_pipeline/profiles/classical-science-zh-Hans。
AGENTS.mdREADME.mdPIPELINE_SPEC.mdautomation_contract.mdtemplate/epub_pipeline/targets/zh-Hans/quality_framework/README.mdreferences/translation_research_universal.mdreferences/quality_standard.mdreferences/ancient_greek_source_notes.mdreferences/ancient_greek_source_types.mdreferences/ancient_greek_textual_criticism_policy.mdreferences/ancient_greek_reference_translation_policy.mdreferences/ancient_greek_names_transliteration_policy.mdreferences/ancient_greek_title_strategy.mdreferences/ancient_greek_to_chinese_translation_notes.mdreferences/stratified_random_spotcheck.mdreferences/release_versioning.mdPreserve source meaning, facts, argument structure, tone, and genre.
保留原文意思、事实、论证结构、语气和文类。
Translate from the Ancient Greek source. Use second-language translations only as reference witnesses.
从古希腊文底本翻译。第二语言译本只能作为参考证据。
When a project uses a facsimile scan, an Ancient Greek transcription, and an English or other second-language translation, keep their roles separate: facsimile for base-image verification, Greek transcription for searchable/source-text control, and translation only for reference.
若项目同时使用扫描本、古希腊文转写和英文等第二语言译本,必须保持三者分工:扫描本用于底本影像核验,古希腊转写用于可检索原文控制,译本只作参考证据。
Write natural Simplified Chinese, not Ancient Greek grammar with Chinese words.
写自然的简体中文,不要写成中文词语套古希腊文语法。
Preserve textual uncertainty. Do not silently “fix” damaged, doubtful, or variant readings.
保留文本不确定性。不得静默修复残损、疑难或异文读法。
Keep names, places, dates, numbers, technical terms, and recurring formulae consistent.
人名、地名、年代、数字、技术术语和反复出现的固定表达必须一致。
Rights and source-edition checks must pass before translation.
版权和底本版本核查通过后才能翻译。
metadata/source_witness_manifest.md must exist before batch translation.
批量翻译前必须存在 metadata/source_witness_manifest.md。
qa/textual/textual_uncertainty_log.md must record variants, conjectures, lacunae, OCR uncertainty, or explicitly state that none were found.
qa/textual/textual_uncertainty_log.md 必须记录异文、拟补、残损、OCR 不确定项,或明确说明未发现。
Book-specific research must identify edition, editor, source structure, textual risks, and reference-witness policy.
本书专项研究必须识别版本、编辑者、来源结构、文本风险和参考译本政策。
Pretranslation report must say PASS before batch translation.
预翻译报告必须 PASS 后才能批量翻译。
Every translated chapter must pass chapter control, fidelity review, readability review, terminology review, and final chapter gate.
每章译文必须通过译后控制、忠实度审校、可读性审校、术语审校和最终章节门禁。
If the classical-science profile is enabled, every relevant chapter must also pass technical and diagram/table audits.
如果启用古典科学 profile,相关章节还必须通过技术审计和图表/表格审计。
After the first full-book EPUB and after every post-EPUB refinement pass, run the stratified random spot-check module: npm run review:random-samples, independent agent review, fix closure, new-seed re-sampling after rework, and npm run review:random-validate:pass before final output. The sampled population is reader-facing audit units, including paragraphs, tables, figures, formulas/proof blocks, captions, and notes.
第一版全书 EPUB 生成后,以及每轮 EPUB 后精校完成后,必须执行分层随机抽检模块:运行 npm run review:random-samples、独立 Agent 评审、修复闭环、返工后新 seed 复抽,并在最终输出前通过 npm run review:random-validate:pass。抽样总体是读者可见审计单元,包括正文段落、表格、图片、公式/证明块、图注和注释。
After random spot-check closure, create a versioned artifact. Public-domain or licensed projects run npm run release:create and write output/release/book_vX.X.X.epub; private_use projects run npm run private:artifact:create and write output/private_artifacts/{title}_private_vX.X.X.epub. output/book.epub alone is not enough for DONE.
随机抽检闭环通过后,必须创建带版本号产物。公版或授权项目运行 npm run release:create 并写入 output/release/book_vX.X.X.epub;private_use 项目运行 npm run private:artifact:create 并写入 output/private_artifacts/{title}_private_vX.X.X.epub。只有 output/book.epub 不能标记 DONE。