gooseworks-ai/goose-skills / skills/video/create-creator-takes-h3
create-creator-takes-h3
VerifiedGenerate an AI creator talking to camera, saying an approved script, as one continuous track — H3 Max reference-to-video through the GooseWorks fal-proxy (charged to your GooseWorks credits). Plans takes on line boundaries under H3's 15s cap, dry-runs a cost estimate, generates the first take alone so its voice can be locked and passed to every later take, joins takes with measured 0.10s dissolves, and moves each line's timing onto the words actually spoken. Use for any format with a generated creator speaking a script (split-screen, screen inserts, talking-head ads).
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 31.6 KB |
| scripts/align_beats.py | script | 3.9 KB |
| scripts/character-prompt.json | config | 2.1 KB |
| scripts/content_review.py | script | 3.3 KB |
| scripts/join_takes.py | script | 12.8 KB |
| scripts/make_character.py | script | 30.0 KB |
| scripts/media_proxy.py | script | 56.8 KB |
| scripts/pair_voice.py | script | 38.1 KB |
| scripts/plan_takes.py | script | 20.7 KB |
| scripts/run_takes.py | script | 19.0 KB |
| scripts/tiers.py | script | 6.1 KB |
| skill.meta.json | config | 278 B |
| tests/smoke-test.md | markdown | 1.3 KB |
| tests/test_action_skin_light.py | script | 7.0 KB |
| tests/test_hands_voice_pace.py | script | 10.1 KB |
| tests/test_make_character.py | script | 4.9 KB |
| tests/test_quality_tier.py | script | 13.5 KB |
| tests/test_run_takes_rejection.py | script | 16.8 KB |
| tests/test_voice_ethnicity.py | script | 34.5 KB |
scripts/content_review.py
View on GitHub"""Read actual source review evidence. Never convert file existence into a pass."""
import hashlib
import json
import pathlib
import subprocess
def source_sha256(path):
digest = hashlib.sha256()
with open(path, "rb") as handle:
for chunk in iter(lambda: handle.read(1 << 20), b""):
digest.update(chunk)
return digest.hexdigest()
def require_take_review(reviews_path, take, path, duration=None):
if not reviews_path:
raise ValueError("supply --reviews with the actual source content review; transcription or file existence is insufficient")
document = json.loads(pathlib.Path(reviews_path).read_text(encoding="utf-8"))
entries = document.get("takes", [])
matches = [entry for entry in entries if entry.get("id") == take["id"]]
if len(matches) != 1:
raise ValueError("take %s needs exactly one content review" % take["id"])
review = matches[0]
if review.get("status") != "pass" or review.get("source_sha256") != source_sha256(path):
raise ValueError("take %s has failed, incomplete or stale source review" % take["id"])
if not take.get("script_sha256") or review.get("script_sha256") != take["script_sha256"]:
raise ValueError("take %s review names a different script" % take["id"])
if not str(review.get("reviewer", "")).strip():
raise ValueError("the actual reviewer is missing")
if duration is None:
duration = float(subprocess.run(["ffprobe", "-v", "error", "-show_entries", "format=duration", "-of", "csv=p=0", str(path)], capture_output=True, text=True, check=True).stdout)
coverage = review.get("coverage", {})
# How the whole take was covered: watched/listened when a person or a capable model really did,
# or measured (full-resolution frames plus word timing against the line). Measured is enough;
# nobody has to claim a watch or a listen. The legacy booleans still mean watched/listened.
motion = coverage.get("motion") or ("watched" if coverage.get("video_watched") is True else None)
audio = coverage.get("audio") or ("listened" if coverage.get("audio_listened") is True else None)
if motion not in ("watched", "measured") or audio not in ("listened", "measured"):
raise ValueError("review must say how the whole take was covered: coverage.motion watched or measured, "
"and coverage.audio listened or measured")
if "measured" in (motion, audio) and not [e for e in review.get("evidence") or [] if str(e).strip()]:
raise ValueError("a measured review lists its evidence: the frames inspected and the word timing checked against the line")
if not (coverage.get("start_s", 1) <= 0 and coverage.get("end_s", 0) >= duration - .02):
raise ValueError("review must cover the complete source")
checks = review.get("checks", {})
for key in ("script", "voice", "lipsync", "identity", "framing", "performance"):
if checks.get(key) != "pass":
raise ValueError("take %s content check %s has not passed" % (take["id"], key))
if checks.get("product_action") not in ("pass", "not_applicable"):
raise ValueError("product/action correctness is unresolved")
if take.get("product_action_required") and checks.get("product_action") != "pass":
raise ValueError("a required product action cannot be marked not_applicable")
return review
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 |