Kling Scene SFX

What problem it solves

Adding scene-matched sound effects to videos, storyboards, or text-based shots usually means searching for assets, syncing timing, and re-rendering repeatedly. This skill sends scene descriptions or video frames to Kling SFX via dLazy's hosted API, producing usable scene SFX for video dubbing, narration atmosphere, or short-form sound drafts.

How it works

• Input: supports text prompts or local image/video/audio file paths; exact parameters are defined by the dlazy keling-sfx command.
• Authentication: requires a dLazy API key, either stored through the CLI or provided via DLAZY_API_KEY.
• Generation: requests are sent to dLazy cloud inference; local media files are uploaded to dLazy media storage so the model can read them, and output URLs are hosted by dLazy.
• Async mode: with --no-wait, the response returns a generateId and status first, then polling is needed with dlazy status instead of receiving final URLs immediately.
• Pipe references: some flags accept -, @N, @N.path, @stdin, etc., to connect upstream output into the current command.

Limits and caveats

• This is a SaaS generation endpoint, not a local-only model; uploaded files should be checked for privacy and asset licensing.
• Common failures include unauthorized, insufficient_balance, missing --prompt, missing local files, or backend safety-policy rejection; handle dLazy dashboard credentials or credits before retrying.
• Output is best treated as a first draft or candidate asset; final usability still depends on visual timing, loudness, reverb, and platform requirements.