Skip to contentVideo input Subtitle recognition Filler clean Import video Detect audio sensitive words Import audio
LmBox
Blueprint lab
Blueprint lab in LmBox is a visual node workflow. Wire video / audio / subtitle input, online or on-device subtitle recognition, rough cut and filler clean, sensitive-word detection, image and sticker generation, OGraf / Lottie graphics, and import into DaVinci Resolve, Premiere and After Effects into one blueprint you can save and run again. You can also turn on external agents in Settings and build and run nodes by talking.
Sensitive-word blocking for voice-over: video input and subtitle recognition, then two paths. One cleans fillers and imports the video; the other detects sensitive words in the audio and imports the audio.
External agents
Connect over MCP
Hand LmBox to an agent such as Cursor or Claude, and build blueprints and run nodes by talking.
Turn on external agents
Go to Settings → Local connections and keep the switch on.
Copy the connection keyword
Click the connection keyword and paste it into the agent chat.
Run a health check first
Once it is connected, let the agent call Blueprint lab.
Ask the agent
Getting started
Nodes
54