Select the best take per script section and generate cropped trim spec paths.
npx skills add https://github.com/theSamPadilla/montaj --skill select-takes
Please help me install this Agent Skill. Skill: select-takes Source: https://github.com/theSamPadilla/montaj/tree/main/skills/select-takes Command: npx skills add https://github.com/theSamPadilla/montaj --skill select-takes
Editing long videos often results in multiple takes of the same script line across clips, wasting runtime and creating redundant content. This Skill automatically analyzes all transcripts, selects the single best take for each script section, and discards the rest, streamlining the editing pipeline.
Tell the agent to execute the select‑takes skill on your Montaj project to receive an ordered list of trimmed spec paths for the best takes.
High-intent search queries and answers about installing and using this skill.
To select the best video take, this skill analyzes cross-clip transcripts and waveform trim specifications to identify the strongest delivery for each script section. It evaluates takes based on completeness, filler count, and recency, discarding redundant clips to streamline the video editing pipeline.
You can automate cropping and trimming specifications by providing SRT files and trim spec JSON files to the skill. It processes these inputs to generate precise keep intervals via the crop_spec API, outputting an ordered list of cropped spec paths for downstream concatenation.
You need to provide SRT files for transcripts and trim spec JSON files for waveform specifications. These inputs are required by the skill to map script sections, select optimal takes, and generate the cropped trim spec files needed for final video assembly.
Duplicate script segments across multiple video angles are handled through cross-clip transcript analysis and seam-overlap checks. The skill selects the single best take for each section, removes in-take repetitions, and eliminates redundant deliveries to produce a clean edit order.
Yes, you can use this skill to generate a highlight reel from multiple camera angles. It automatically selects the optimal take for each script segment across all clips, producing an ordered list of trimmed spec paths for downstream filler removal and clean video concatenation.
Discarded video takes are bypassed during the generation of cropped trim spec files. The skill focuses solely on the selected optimal takes, creating precise keep intervals that ensure downstream filler removal and concatenation only process the chosen video segments.
theSamPadillaGenerate waveform images and output crop_spec JSON for silence detection.
GeneralReasoningConvert filler-word timestamp annotations into merged cut segments for video processing.
edwifiguyEdit screen recordings programmatically via a stateful CLI with ffmpeg.
video-production-buddyAutomate video and audio processing tasks with FFmpeg commands.
