camera-videolisted

# Camera Video This skill executes video workflows and does not author or audit prompts. Supply the model-native H3 prompt directly: ```json {"prompt": {"text": "..."}} ``` The envelope contains only `prompt.text`. Use `duration` and the required local reference-image paths in `config`; `describe_config` lists stage-specific requirements. Call `describe_config`, then `validate_config`, then `run_skill`.