russian-lyrics-romanizationlisted
Install: claude install-skill XXXG00W0/russian-lyrics-romanization
# Russian lyrics → romanization + English pronunciation
Produce three aligned lines for each line of lyrics:
```
Я не могу́ забы́ть твои́ глаза́ original, with stress marks
Ya ne mogu zabyt tvoi glaza romanization (letter-level transliteration)
ya nye ma-GOO za-BIT tva-EE gla-ZA English pronunciation (stress in CAPITALS)
```
The romanization line is a letter-level transliteration. The pronunciation line is
phonetic: it applies the pronunciation rules (unstressed о is spoken "a", final
consonants devoice, -ться is "-tsa").
## Do not hand-write the output
The romanization and the pronunciation line are produced by the rule engine.
Write the text to a file (or pipe it) and run the scripts. Hand-transliterating a
song drifts from the rules between lines and can break the word-for-word alignment.
Stress in context is the part that needs a reader - see "Resolving stress" below.
## Workflow
```
1. save the lyrics -> song.txt (strip LRC timestamps and metadata first)
2. stress.py -> cached stress DB lookup; returns marked text + report
3. resolve the report -> pick homograph readings, handle unknown words
4. render.py -> the three lines
5. verify.py --check -> sanity check (optional but cheap)
6. present -> tell the user which words were inferred, not looked up
```
### 1. Get the text
Lyrics may arrive pasted, as a file, or as an `.lrc` with `[00:12.34]` timestamps and
`[ar:...]`/`[ti:...]` tags. Anyth