Dr Cooken To Zh

Workflow

Step 1 — Check upstream context. Inspect context_output. If context_output.raw_text exists and context_output.language is en, use that text as the source. Otherwise prompt the user: "Please paste the English text you would like translated to Chinese."

Step 2 — Collect missing required parameters. If domain is not present in either the user's request or context_output.parameters, ask once: "What is the domain of this text? (tcm / bioinformatics / clinical / pharmacology)". Wait for the answer before proceeding. Do not solicit optional parameters unless the user volunteers them.

Step 3 — Load term-mapping.md. Read references/term-mapping.md. For domain: tcm, use the TCM Core Concepts table for reverse lookup (English → Chinese). For domain: bioinformatics, use the Bioinformatics Terms table to select established Chinese equivalents. For clinical and pharmacology domains, rely on discipline-specific international nomenclature adapted to Chinese academic conventions (NMPA terminology for pharmacology; national clinical reporting standards for clinical). Keep the loaded tables active through Steps 4 and 5.

Step 4 — Translate the full text.

Apply the following rules:

• Formal written Chinese (书面语): Use formal academic Chinese throughout. Avoid spoken register (口语), colloquialisms, or overly casual phrasing. Sentence structures should reflect Chinese academic norms, not a direct transposition of English syntax.
• Sentence restructuring: English academic sentences often use extensive subordinate clauses and relative clauses that become unwieldy in direct Chinese translation. Restructure to natural Chinese academic flow: Chinese academic sentences can be longer and may use parallel structures (排比) effectively; avoid over-fragmentation of English compound sentences.
• Bioinformatics terms (domain: bioinformatics): Use established Chinese equivalents from term-mapping.md as primary terms. On first use of each term, append the English abbreviation in parentheses: "差异表达基因 (DEG)", "基因集富集分析 (GSEA)". Subsequent occurrences may use either the Chinese term or the abbreviation alone, whichever is clearer in context.
• Gene and protein names: Retain English HGNC names exactly as-is. Do not translate gene symbols (TP53 remains TP53; NF-κB remains NF-κB including Greek letters and formatting). Gene names are not italicized in Chinese academic convention but retain their exact English form inline.
• TCM terminology (domain: tcm, preserve_tcm_terms: true): Use GB/T standard Chinese terms (GB/T 16751.1-1997 where applicable). If translating from English TCM equivalents (e.g., "syndrome differentiation and treatment"), look up the canonical Chinese form via term-mapping.md reverse lookup (辨证论治). Do not invent new Chinese TCM terms for concepts that have established equivalents.
• Units and statistics: Keep SI units, p-values, confidence intervals, and numerical values exactly as written in the source. Do not translate unit symbols or reformat statistical expressions.

Step 5 — Post-translation quality check.

Scan the draft translation for:

• Literal translations that sound unnatural in academic Chinese: e.g., avoid "在…的情况下" where a more direct construction is possible; avoid translating every English connector (however, furthermore) with its most literal Chinese counterpart when the Chinese sentence structure makes them redundant.
• Sentence length balance: check that English complex sentences have been restructured rather than simply rendered as a single overlong Chinese sentence; and that English short sentences have not been over-fragmented into staccato Chinese clauses.
• TCM term standards: verify that TCM terms match national standard equivalents from GB/T 16751.1-1997 as reflected in term-mapping.md. Flag any term where only an informal or non-standard Chinese form was available.
• English gene/protein names: confirm all HGNC symbols are retained in their original English form and are not altered or translated.

Step 6 — Present output.

Deliver the translated text, then a brief glossary of key technical terms handled during translation. Limit the glossary to 10 entries; choose entries where a non-obvious translation decision was made (e.g., a bioinformatics abbreviation with multiple possible Chinese renderings, or a TCM term requiring reverse lookup). Do not list routine vocabulary.

Step 7 — Offer next step.

After presenting output, ask: "Would you like me to polish this Chinese text for academic style? (runs chinese-polisher)"

Output Format

## 译文

[Full Chinese text]

---
## 关键术语处理

| 英文原文 | 中文译文 | 说明 |
|---------|---------|------|
| DEG | 差异表达基因 (DEG) | 保留英文缩写 |
| syndrome differentiation | 辨证论治 | 按GB/T标准 |
| TP53 | TP53 | 保留英文HGNC格式 |

If any translation choices were flagged in Step 5 as potentially non-standard, append:

---
## 术语说明

- <term> — <reason this choice was made or alternative to consider>

context_output

Reads from upstream:

• raw_text — source English text (if passed from a prior module)
• domain — domain override from upstream
• target_journal — journal context from upstream
• language — used to confirm source language is en

Writes to downstream:

{
  "module": "en-to-zh",
  "domain": "<collected>",
  "language": "zh",
  "raw_text": "<Chinese translation>",
  "summary": "Translated [N words] from English to Chinese; [N] technical terms standardized",
  "parameters": {
    "target_audience": "academic_journal",
    "preserve_tcm_terms": true
  },
  "status": "success",
  "error_message": "string | null"
}

References

• references/term-mapping.md — canonical English↔Chinese term tables for TCM and bioinformatics domains. Loaded at Step 3.