Transcribe a recorded audio file (meeting / call / voice memo / interview) with the local whisper tool, then summarize the transcript into a readable note with key points, steered by the focus variable. Content-isolated: audio and transcript move as Gateway-gated file refs, never through step values. Preview-first: shows the note and writes nothing; set write_mode to stage to persist it (with a transcript pointer) into data.shared. Source: a local audio path or a data.file ref from another step.

How it works 15 steps

Inspect the data fetches, transforms, gates, and output this recipe runs.

Process (15 steps)

■ transcribed ?

skip: config.source is_empty

■ file_ref default

Apply default

■ has_file compare

Check if file ref has data

■ summarize ?

skip: step.has_file not_equal true

■ summary default

Apply default

■ key_points default

Apply default

■ key_points_count count

Count items in key points

■ key_points_text join

Join items with " "

■ now date_add

Add 0 days to

■ skip_write any

Check if any of the conditions are true

■ keep ?

skip: step.skip_write equal true

■ row_key replace

Replace text in keep cas ref

■ stage_row ?

skip: step.skip_write equal true

■ staged not

Apply not

■ card to_summary

Format results as a summary card

Settings 6 configurable

Configurable at install. Defaults shown — change them anytime in Recued.

focus setting = key decisions, action items, and open questions

title setting =

source setting =

max length setting = 250

write mode setting = preview

staging prefix setting = media_transcripts_staging