pronunciation-specialist
정보
이 스킬은 Suno AI 음악 생성에서의 잘못된 발음을 방지하기 위해, 가사 속 고유명사, 전문 용어, 비영어 단어와 같은 발음 위험 요소를 검사합니다. 제공된 가사를 분석하거나 트랙 파일(기악 부분 제외)을 읽어 문제점과 수정 제안을 보고합니다. 자동 보컬 합성이 발음을 잘못할 수 있는 요소가 포함된 가사를 작성할 때 사용하세요.
빠른 설치
Claude Code
추천npx skills add bitwize-music-studio/claude-ai-music-skills -a claude-code/plugin add https://github.com/bitwize-music-studio/claude-ai-music-skillsgit clone https://github.com/bitwize-music-studio/claude-ai-music-skills.git ~/.claude/skills/pronunciation-specialistClaude Code에서 이 명령을 복사하여 붙여넣어 스킬을 설치하세요
문서
Your Task
Input: $ARGUMENTS
Instrumental Guard
When invoked with a track file path, first check the track's frontmatter for instrumental: true or the Track Details table for **Instrumental** | Yes. If the track is instrumental:
- STOP and report: "SKIP — Instrumental track (no lyrics to scan for pronunciation)"
- Do NOT scan instrumental tracks.
Vocal Track Workflow
Based on the argument provided:
- If given a track file path: Read it, scan lyrics for pronunciation risks, report issues with fixes
- If given lyrics directly: Scan and flag risky words
- Output: Clean lyrics with all phonetic fixes applied, ready for suno-engineer
Supporting Files
- word-lists.md - Complete tables of homographs, tech terms, names, acronyms, numbers
Pronunciation Specialist
Scan lyrics for pronunciation risks, suggest phonetic spellings, prevent Suno mispronunciations.
Why This Matters
The problem: Suno AI guesses pronunciation. Wrong guess = wrong song = wasted generation.
One wrong word ruins the take.
When to Invoke
Always invoke between lyric-writer and lyric-reviewer:
lyric-writer (WRITES + SUNO PROMPT) → pronunciation-specialist (RESOLVES) → lyric-reviewer (VERIFIES) → pre-generation-check
|
Scan, resolve, fix risky words
Your role — RESOLVE:
- The lyric-writer flags potential pronunciation risks and asks about homographs
- You do the deep scan, resolve ambiguities with the user, and apply all phonetic fixes
- The lyric-reviewer then verifies all resolutions were correctly applied
High-Risk Word Categories
See word-lists.md for complete tables. Summary:
1. Homographs (CRITICAL)
Same spelling, different pronunciation. ALWAYS require clarification.
(Canonical reference: ${CLAUDE_PLUGIN_ROOT}/reference/suno/pronunciation-guide.md. Keep this summary in sync.)
| Word | Options | Fix |
|---|---|---|
| live | LYVE (verb) / LIV (adjective) | "lyve" or "liv" |
| read | REED (present) / RED (past) | "reed" or "red" |
| lead | LEED (guide) / LED (metal) | "leed" or "led" |
| wind | WYND (air) / WINED (coil) | "wynd" or "wined" |
| tear | TEER (cry) / TARE (rip) | "teer" or "tare" |
| bass | BAYSS (music) / BASS (fish) | "bayss" or "bass" |
2. Tech Terms
Suno often mispronounces tech words:
- Linux → "Lin-ucks" (not "Line-ucks")
- SQL → "S-Q-L" or "sequel"
- API, CLI, SSH → spell out with hyphens
3. Names & Proper Nouns
Non-English names need phonetic spelling:
- Jose → "Ho-zay"
- Ramos → "Rah-mohs"
- Sinaloa → "Sin-ah-lo-ah"
4. Acronyms
3-letter acronyms → spell out with hyphens (FBI → F-B-I) Word-like acronyms → phonetic (RICO → Ree-koh, NASA → Nah-sah)
5. Numbers
- Years: Use apostrophes ('93) or words (nineteen ninety-three)
- Digits: Write out (four-oh-four, not 404)
Pronunciation Guides
You reference TWO pronunciation guides:
Base Guide (Plugin-Maintained)
- Location:
${CLAUDE_PLUGIN_ROOT}/reference/suno/pronunciation-guide.md - Contains: Universal pronunciation rules, common homographs, tech terms
- Updated: By plugin maintainers when new issues are discovered
Override Support
Check for custom pronunciation entries:
Loading Override
- Call
load_override("pronunciation-guide.md")— returns override content if found (auto-resolves path from config) - If found: load and merge with base guide (override entries take precedence)
- If not found: use base guide only (skip silently)
Override File Format
{overrides}/pronunciation-guide.md:
# Pronunciation Guide (Override)
## Artist Names
| Name | Pronunciation | Notes |
|------|---------------|-------|
| Ramos | Rah-mohs | Character name |
## Album-Specific Terms
| Term | Pronunciation | Notes |
|------|---------------|-------|
| Sinaloa | Sin-ah-lo-ah | Location |
How to Use Override
- Add artist names, album-specific terms, and genre-specific jargon
- Override entries take precedence over base guide entries for the same word
- Base guide updates via plugin updates without conflicts
- Override guide is version-controlled with your music content
Scanning Workflow
Step 1: Automated Scan via MCP
- Extract lyrics:
extract_section(album_slug, track_slug, "lyrics") - Homograph scan:
check_homographs(lyrics_text)— returns found homographs with line numbers, pronunciation options - Additional manual scan for tech terms, acronyms, numbers, and names (not covered by MCP homograph list) — cross-reference word-lists.md
- If style prompt exists:
scan_artist_names(style_text)— catch blocklisted names
After fixes are applied:
5. Verify: check_pronunciation_enforcement(album_slug, track_slug) — confirms all pronunciation table entries appear in lyrics
Step 2: Review Results
From MCP results and manual scan:
- Which words were flagged?
- What's the recommended fix for each?
Step 3: Generate Report
For each flagged word, provide:
- Line number and context
- Why it's risky (ambiguity type)
- Suggested phonetic spelling
- Alternative if multiple pronunciations exist
Example output:
PRONUNCIATION RISKS FOUND (3):
Line V1:3 -> "We live in darknet spaces"
Risk: "live" is homograph
Options: "lyve" (verb) or "liv" (adjective)
-> Needs clarification
Line C:1 -> "SQL injection in the code"
Risk: "SQL" is tech acronym
Fix: "S-Q-L" or "sequel"
-> Auto-fix: "S-Q-L injection in the code"
Line V2:5 -> "Reading Linux logs at 3AM"
Risk: "Linux" commonly mispronounced
Fix: "Lin-ucks"
-> Auto-fix: "Reading Lin-ucks logs at 3 A-M"
Step 4: User Confirmation
For ambiguous words (like "live"): Ask user which pronunciation For clear fixes (tech terms): Auto-fix
Auto-Fix Rules
Always Auto-Fix
- Tech terms (SQL → S-Q-L, Linux → Lin-ucks)
- Common acronyms (FBI → F-B-I, GPS → G-P-S)
- Numbers (1993 → '93 or nineteen ninety-three)
Ask User First
- Homographs (live, read, lead, wind, tear)
- Names (confirm pronunciation preference)
- Words with regional variants (data, either, route)
Output Format
Track File Updates
If given a track file, update these sections:
Pronunciation Notes (add table):
| Word/Phrase | Phonetic | Notes |
|-------------|----------|-------|
| Jose Diaz | Ho-say Dee-ahz | Spanish name |
| live | lyve | Verb form (to reside) |
| SQL | S-Q-L | Spell out |
Lyrics Box (apply fixes): Replace standard spelling with phonetic in the Suno lyrics section.
Standalone Report
PRONUNCIATION SCAN COMPLETE
===========================
File: [path or "direct input"]
Risks found: X
Auto-fixed: Y
Needs user input: Z
FIXES APPLIED:
- "SQL" → "S-Q-L" (line V1:3)
- "Linux" → "Lin-ucks" (line V2:5)
NEEDS USER INPUT:
- "live" (line C:1) - lyve or liv?
CLEAN LYRICS:
[Full lyrics with all fixes applied]
Adding Custom Pronunciations
When you discover new pronunciation issues specific to the user's content:
Add to OVERRIDE guide ({overrides}/pronunciation-guide.md):
- Read config to get
paths.overrideslocation - Check for
{overrides}/pronunciation-guide.md - Create file if it doesn't exist (with header and table structure)
- Add the word to appropriate section (Artist Terms, Album Names, etc.)
- Include: word, standard spelling, phonetic spelling, notes
Example entry:
| Larocca | larocca | Luh-rock-uh | Character in "sample-album" album |
DO NOT edit the base guide (${CLAUDE_PLUGIN_ROOT}/reference/suno/pronunciation-guide.md) - plugin updates will overwrite it.
When to add:
- Artist names, album titles, track titles
- Character names in documentary/narrative albums
- Location names specific to album content
- Any pronunciation discovered during production
This keeps discoveries version-controlled with the music content in the overrides directory.
Remember
- Load both guides at start - Base guide + override guide (if exists)
- Homographs are landmines - live, read, lead, wind WILL mispronounce without fixes
- Tech terms need phonetic spelling - Don't trust Suno with acronyms
- Non-English names always need help - Phonetic spelling mandatory
- Numbers are tricky - Write them out or use apostrophes
- When in doubt, ask - Better to clarify than regenerate
- Add discoveries to OVERRIDE guide - Never edit base guide (plugin will overwrite)
GitHub 저장소
연관 스킬
release-standards
문서 처리이 스킬은 소프트웨어 릴리스에 대한 시맨틱 버저닝(semver) 가이드라인과 변경 로그 형식 표준을 제공합니다. 릴리스를 준비할 때 버전 번호(메이저/마이너/패치)를 올바르게 증가시키고 변경 로그 항목을 구성하려면 이 스킬을 사용하세요. 사전 릴리스 식별자 규칙과 개발자를 위한 명확한 예시가 포함되어 있습니다.
commit-standards
문서 처리이 스킬은 Conventional Commits 표준에 따라 Git 커밋 메시지를 형식화합니다. 커밋 작성이나 리뷰 시 일관성을 보장하기 위해 템플릿과 유형 정의(예: `feat`, `fix`, `refactor`)를 제공합니다. 커밋 과정에서 이를 사용하여 명확하고 구조화된 커밋 기록을 생성할 수 있습니다.
huggingface-tokenizers
문서 처리이 스킬은 HuggingFace의 Rust 기반 라이브러리를 사용하여 1GB 텍스트를 20초 이내에 처리하는 고성능 토크나이제이션을 제공합니다. BPE, WordPiece, Unigram 알고리즘을 지원하며 사용자 정의 토크나이저 학습과 정렬 추적 기능을 포함합니다. 프로덕션 수준의 고속 토크나이제이션이 필요하거나 transformers 생태계와 통합된 맞춤형 토크나이저를 구축할 때 사용하세요.
nano-pdf
문서 처리nano-pdf는 개발자가 특정 페이지의 텍스트 변경이나 오타 수정과 같은 자연어 지시를 사용해 PDF를 편집할 수 있는 CLI 도구입니다. 터미널에서 직접 빠르고 프로그래밍 방식으로 PDF를 수정하는 데 이상적입니다. 페이지 번호 매기기가 버전마다 다를 수 있으므로 출력 결과는 항상 확인하세요.
