puntorigen/avatar-skills / seedance-2
seedance-2
UnverifiedGenerate video with ByteDance Seedance 2.0 via the Higgsfield MCP (generate_video, model seedance_2_0). A unified multimodal model that turns a text prompt plus reference images, video clips, and audio into audio-synced video up to 15s, with strong physics, multi-shot camera planning, and lip-sync. Use to generate video from text, animate a still image (image-to-video), do motion transfer / character consistency / style reference / audio-driven generation from references, or animate a storyboard sheet into a movie. Reframes prompts with [Image1]/[Video1]/[Audio1] tokens for references. Trigger when the user mentions seedance, seedance-2, seedance 2.0, generating a video (of a given duration/resolution/ratio), animating a storyboard, or video with image/video/audio references.
| File | Kind | Size |
|---|---|---|
| REFERENCE.md | markdown | 7.7 KB |
| SKILL.md | skill md | 13.8 KB |
| prompts/storyboard_video_framework.md | markdown | 7.2 KB |
| scripts/prompt_tools.py | script | 5.9 KB |
scripts/prompt_tools.py
View on GitHub#!/usr/bin/env python3
"""Prompt helpers for the seedance-2 skill (Higgsfield MCP backend).
Generation itself runs through the Higgsfield MCP `generate_video` tool
(model `seedance_2_0`), which the agent calls directly. This script is a
pure-text helper (no network, no deps) that guarantees the exact prompts:
storyboard -> the SIMPLE one-liner storyboard->video prompt (FALLBACK). Use
this only when the storyboard was NOT built via the guide's
Phase 1. If it was, author the Phase 2 per-shot cinematic video
prompt instead, following prompts/storyboard_video_framework.md.
reframe -> wire reference tokens ([Image1]/[Video1]/[Audio1]) into a prompt
Examples:
python3 prompt_tools.py storyboard
python3 prompt_tools.py storyboard --panels 4-6
python3 prompt_tools.py reframe "A hero walks in" --images 2 --videos 1 \
--audios 1 --audio-transcript "It's not just a pretzel, Arthur!"
"""
import argparse
import re
import sys
# The storyboard guide's fixed audio line, kept verbatim (em-dash, capital "Diegetic").
AUDIO_LINE = ("Audio: Diegetic sound only — natural ambience, environmental foley, "
"and subject-driven sound.")
def parse_panel_range(spec):
"""Parse '4-6', '4 to 6', '4', '4,6' into (start, end|None); None -> whole board."""
if spec is None:
return None
nums = [int(n) for n in re.findall(r"\d+", str(spec))]
if not nums:
return None
if len(nums) == 1:
return (nums[0], None)
return (nums[0], nums[1])
def build_storyboard_prompt(panel_range, extra=None):
if panel_range is None:
scope = "from panels"
else:
start, end = panel_range
scope = f"from panel {start}" if end is None else f"from panels {start} to {end}"
prompt = (f"Use the reference storyboard to make a full animation movie {scope}. "
f"{AUDIO_LINE}")
if extra:
prompt += f" {extra.strip()}"
return prompt
def reframe_prompt_with_references(prompt, n_images, n_videos, n_audios,
audio_transcript=None):
"""Ensure each provided reference asset is wired into the prompt by token.
Seedance 2.0 references assets via tokens: [Image1], [Video1], [Audio1],
mapped in order to the same-type medias passed to generate_video. Tokens
already present in the prompt are left untouched; missing ones get a
neutral reference clause appended. Prefer writing tokens yourself for
precise semantic control (e.g. "[Image2] stands in the room of [Image1]").
"""
clauses = []
missing_imgs = [f"[Image{i}]" for i in range(1, n_images + 1)
if f"[Image{i}]" not in prompt]
if missing_imgs:
clauses.append(
f"Use {', '.join(missing_imgs)} as references for character appearance, "
f"style, and composition.")
missing_vids = [f"[Video{i}]" for i in range(1, n_videos + 1)
if f"[Video{i}]" not in prompt]
if missing_vids:
clauses.append(
f"Use {', '.join(missing_vids)} as references for motion and camera style.")
missing_auds = [f"[Audio{i}]" for i in range(1, n_audios + 1)
if f"[Audio{i}]" not in prompt]
if missing_auds:
if audio_transcript:
clauses.append(f'The subject says {missing_auds[0]}: "{audio_transcript.strip()}".')
else:
clauses.append(f"Sync the dialogue/audio from {', '.join(missing_auds)}.")
if not clauses:
return prompt
return prompt.rstrip() + "\n\n" + " ".join(clauses)
def main():
parser = argparse.ArgumentParser(
description="Prompt builders for the seedance-2 (Higgsfield) skill.",
formatter_class=argparse.RawDescriptionHelpFormatter,
)
sub = parser.add_subparsers(dest="command", required=True)
sb = sub.add_parser(
"storyboard",
help="Build the SIMPLE one-liner storyboard->video prompt (fallback; author "
"the Phase 2 cinematic prompt instead when the board came from Phase 1)")
sb.add_argument(
"--panels", "-p", default=None,
help="Panel range, e.g. '4-6', '4 to 6', or a single '4'. Omit for the full board.")
sb.add_argument("--extra", default=None,
help="Optional extra direction appended after the one-liner")
rf = sub.add_parser("reframe", help="Wire reference tokens into a prompt")
rf.add_argument("prompt", nargs="?", help="Base prompt text (or use --prompt-file)")
rf.add_argument("--prompt-file", help="Read the base prompt from a file")
rf.add_argument("--images", type=int, default=0, help="Number of reference images")
rf.add_argument("--videos", type=int, default=0, help="Number of reference videos")
rf.add_argument("--audios", type=int, default=0, help="Number of reference audios")
rf.add_argument("--audio-transcript", help="Spoken line woven in for lip-sync")
args = parser.parse_args()
if args.command == "storyboard":
print(build_storyboard_prompt(parse_panel_range(args.panels), extra=args.extra))
return
if args.command == "reframe":
if args.prompt_file:
import pathlib
pf = pathlib.Path(args.prompt_file)
if not pf.exists():
parser.error(f"--prompt-file not found: {pf}")
prompt = pf.read_text(encoding="utf-8").strip()
elif args.prompt:
prompt = args.prompt
else:
parser.error("Provide a prompt or --prompt-file")
if args.audios and not (args.images or args.videos):
print("Note: seedance audio references require at least one image or video reference.",
file=sys.stderr)
print(reframe_prompt_with_references(
prompt, args.images, args.videos, args.audios,
audio_transcript=args.audio_transcript))
return
if __name__ == "__main__":
main()
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |