| name | chapter-intro |
| description | Write chapter introductions for biblical text chapters summarizing themes and key content. Use when asked to write a chapter introduction or after pipeline completes. |
| allowed-tools | Read, Grep, Glob, Write, Edit |
Chapter Introduction
Generate a brief chapter introduction that orients translators to the chapter's overall movement, key concepts, and any distinctive translation challenges. Target length: 300-600 characters. Run this after the pipeline has completed for a chapter.
Arguments
When invoked as /chapter-intro isa 51:
- First argument: Book abbreviation (psa, gen, 2sa, etc.)
- Second argument: Chapter number
- Optional hint args from notes pipeline:
--parallelism-signal high--parallelism-count <N>
Book abbreviations follow standard 3-letter codes or common variants:
- psa, ps -> PSA
- 2sam, 2sa -> 2SA
- gen -> GEN
Pipeline Context
If --context <path> is provided, read the context.json file for authoritative source paths (sources.ult, sources.ust, sources.issues, sources.hebrew). Use these instead of searching for files.
If artifacts.parallelism_signal is present in context.json (or --parallelism-signal high is passed), treat it as a chapter-level hint only: mention recurring parallelism briefly in "Translation Issues in This Chapter" if that section is included, without adding verse-level detail.
Workflow
Step 1: Gather Inputs
Normalize arguments:
- Book code: uppercase 3-letter code (e.g.,
PSA)
- Chapter: zero-padded 3-digit for filenames (e.g.,
061), plain number for references (e.g., 61)
Read the following files. Not all may exist; work with what's available.
Pipeline output (preferred):
- ULT:
output/AI-ULT/<BOOK>/<BOOK>-<CHAPTER_PAD>.usfm
- Issues:
output/issues/<BOOK>/<BOOK>-<CHAPTER_PAD>.tsv
- UST:
output/AI-UST/<BOOK>/<BOOK>-<CHAPTER_PAD>.usfm
Fallbacks if pipeline output doesn't exist:
- Published ULT:
data/published_ult_english/ (find the book file, extract the chapter)
- Published UST:
data/published_ust_english/ (find the book file, extract the chapter)
Always read:
- Hebrew source: find the chapter in
data/hebrew_bible/ (search for the book's USFM file containing the chapter)
Reference materials (consult as needed):
- Published TN intros for style reference: published TN intro rows for the same book when available
- Translation Words for key terms: use
mcp__workspace-tools__check_tw_headwords with terms=["term1", "term2"] or browse data/en_tw/
- Translation Academy articles relevant to the chapter's literary form or recurring translation issues
Step 2: Read the Style Guide
Read .claude/skills/reference/gl_guidelines.md for spelling, punctuation, and register rules that apply to all generated content (American English spelling, Oxford comma, curly quotes, formal register, etc.).
Step 3: Determine Chapter Function
Identify the chapter's role in the book and its dominant literary movement. Look for:
- speaker changes
- shifts between narration, poetry, prophecy, exhortation, or prayer
- repeated themes or key images
- distinctive translation challenges that affect the whole chapter rather than one verse
This feeds the 1-2 sentence overview in the Structure and Formatting section.
Step 4: Draft the Introduction
Write a short intro using the template below. The entire intro should be 300-600 characters. Each section is 1-2 sentences. Only include the Translation Issues section when there is a genuinely distinctive challenge (speaker ambiguity, extended metaphor spanning multiple verses, abrupt shifts in audience, repeated legal or ritual terms, etc.).
When a high parallelism hint is present, keep the note brief (one short sentence) and chapter-level (for example, recurring synonymous parallel lines), without listing individual verses.
Use [[rc://*/tw/dict/bible/kt/<term>]] or [[rc://*/tw/dict/bible/other/<term>]] for Translation Word links. Use [Book Chapter](../book/chapter/verse.md) for cross-references.
Template
# <Book> <Chapter> Introduction
## Structure and Formatting
[1-2 sentences: chapter function, literary movement, and brief characterization. Include a tW or tA link when it materially helps translators.]
## Religious and Cultural Concepts in This Chapter
### [Concept Name]
[1-2 sentences explaining the concept.]
### [Optional second concept]
[1-2 sentences.]
## Translation Issues in This Chapter (optional)
### [Issue Name]
[1-2 sentences. Only include for distinctive challenges.]
Quality Checks
- Introduction matches the chapter's actual content and literary function
- tW links use correct
[[rc://...]] format
- Content is translator-oriented, not devotional
- No verse-level detail that belongs in translation notes
- Total length stays in the 300-600 character range
Step 5: Format and Insert into Issue File
Format the intro for TSV storage:
- Escape actual newline characters as literal
\n (two characters: backslash + n). This is the ONLY escaping needed. Do NOT escape Unicode characters — en-dashes (–), em-dashes (—), curly quotes (“ ” ‘ ’), apostrophes (’), or any other non-ASCII character — as \u sequences. Write all such characters as their actual Unicode characters.
- The intro content goes in column 7 (the explanation/content column)
Build the issue TSV row (7 tab-separated columns, matching issue TSV format):
<book>\t<chapter>:intro\t\t\t\t\t<escaped intro content>
Example:
hab 3:intro # Habakkuk 3 Introduction\n\n## Structure and Formatting\n\n...
Columns:
- Book code (lowercase, e.g.,
psa)
- Reference:
<chapter>:intro
- Issue type: (empty)
- ULT quote: (empty)
- Go? flag: (empty)
- AT: (empty)
- Content: the full intro with
\n escapes
Insert into the issue file:
- Read the existing issue file at
output/issues/<BOOK>/<BOOK>-<CHAPTER_PAD>.tsv
- Prepend the intro row as the first line
- Write back to the same file
If the issue file already has an intro row (first line contains :intro), replace it rather than adding a duplicate.
Confirm the result by reading back the first 3 lines of the file to verify:
- The intro row is correctly placed as the first data line
- The intro is a single TSV line — no actual newline characters in the content. All markdown line breaks must be literal two-character
\n sequences. If you see the intro spanning multiple lines in the file, fix it immediately.
- There are no
\u escape sequences in the content (e.g., \u2013, \u201c). If you see any, the file must be rewritten with the actual Unicode characters substituted in place of the escape sequences.
