gooseworks-ai/goose-skills / skills/video/create-vo-elevenlabs
create-vo-elevenlabs
VerifiedGenerate a voiceover (VO) clip via ElevenLabs text-to-speech, routed through the GooseWorks proxy, so the call is charged to your GooseWorks credits. Voice id + script text come from the template recipe. Use for the spoken narration of VO-driven video-ad formats (cgi-app-sizzle, flat-vector-explainer, hypermotion). Never call ElevenLabs directly — the proxy attribution is required.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 5.9 KB |
| scripts/gen_vo.py | script | 12.1 KB |
| scripts/media_proxy.py | script | 56.8 KB |
| scripts/read_pronunciations.py | script | 3.7 KB |
| skill.meta.json | config | 221 B |
| tests/smoke-test.md | markdown | 98 B |
| tests/test_gen_vo.py | script | 1.0 KB |
| tests/test_pace.py | script | 4.1 KB |
| tests/test_timestamps.py | script | 3.3 KB |
scripts/read_pronunciations.py
View on GitHub#!/usr/bin/env python3
"""Export confirmed pronunciation facts from a fresh brand-context read, not a new store.
A confirmed pronunciation is a user-authored `must` brand fact. The canonical text is
Pronounce "<written term>" as "<spoken form>"
(the form the goose-video entry skill saves). The older
Pronunciation: <written term> => <spoken form>
is still read. The context may be the brand read's payload, the MCP structured result
`{"result": {...}}`, or the whole tool result with `structuredContent`/`content`."""
import argparse
import json
import pathlib
import re
_Q = "\"“”" # straight and curly double quotes
FORMATS = (
re.compile(r"Pronounce\s+[%s](.+?)[%s]\s+as\s+[%s](.+?)[%s]\.?" % (_Q, _Q, _Q, _Q), re.IGNORECASE),
re.compile(r"Pronunciation:\s*(.+?)\s*=>\s*(.+)"),
)
def parse_fact(text):
"""Return (term, say_as) for a pronunciation fact's text, or None."""
text = (text or "").strip()
for pattern in FORMATS:
m = pattern.fullmatch(text)
if m:
term, say = (x.strip() for x in m.groups())
if term and say:
return term, say
return None
def unwrap(context):
"""Accept the brand payload, `{"result": payload}`, or a whole MCP tool result."""
for _ in range(3):
if not isinstance(context, dict) or "learnings" in context or "brand" in context:
break
if isinstance(context.get("structuredContent"), dict):
context = context["structuredContent"]
elif isinstance(context.get("result"), dict):
context = context["result"]
elif isinstance(context.get("content"), list):
texts = [c.get("text") for c in context["content"] if isinstance(c, dict) and c.get("type") == "text"]
if len(texts) != 1:
break
context = json.loads(texts[0])
else:
break
if not isinstance(context, dict):
raise ValueError("brand context is not a JSON object")
return context
def rules_from_context(context, brand_id, required=()):
context = unwrap(context)
if (context.get("brand") or {}).get("id") != brand_id:
raise ValueError("brand context does not match the selected brand")
if "learnings" not in context:
raise ValueError("read the brand's learnings before planning voice")
found = {}
for fact in context["learnings"]:
if fact.get("source") != "user" or fact.get("kind") != "must" or fact.get("active") is False:
continue
parsed = parse_fact(fact.get("text"))
if not parsed:
continue
term, say = parsed
if term.lower() in found and found[term.lower()]["say_as"] != say:
raise ValueError("conflicting confirmed pronunciations for " + term)
found[term.lower()] = {"term": term, "say_as": say, "fact_id": fact["id"]}
missing = [term for term in required if term.lower() not in found]
if missing:
raise ValueError("confirm and persist pronunciation before voice: " + ", ".join(missing))
return {"brand_id": brand_id, "basis": "fresh user-authored brand learnings read",
"pronunciations": list(found.values())}
def main():
ap=argparse.ArgumentParser(description=__doc__, formatter_class=argparse.RawDescriptionHelpFormatter)
ap.add_argument("--context", required=True)
ap.add_argument("--brand", required=True)
ap.add_argument("--require", action="append", default=[])
ap.add_argument("--out", required=True)
a=ap.parse_args()
result=rules_from_context(json.loads(pathlib.Path(a.context).read_text()),a.brand,a.require)
pathlib.Path(a.out).write_text(json.dumps(result,indent=2))
print("verified pronunciation rules ->",a.out)
if __name__ == "__main__":
main()
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| slack-gif-creatoranthropics/skillsKnowledge and utilities for creating animated GIFs optimized for Slack. Provides constraints, validation tools, and animation concepts. Use when users request animated GIFs for Slack like "make me a GIF of X doing Y for Slack." | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| ponytailDietrichGebert/ponytailLazy senior dev mode for any coding task: the smallest change that fully solves it, and a reply a busy human understands in one read. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailHarvest every shortcut: comment into one debt ledger, so deferrals get tracked instead of forgotten. One-shot report. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-debtDietrichGebert/ponytailList every `shortcut:` comment (and older `ponytail:` ones) as a debt ledger. One-shot report, changes nothing. Use for "ponytail debt", "what did ponytail defer", "list the shortcuts", /ponytail-debt. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |