Use “Transcribe audio to commitments” in Recued. Transcribe a recorded audio file (meeting / call / standup) with the local whisper tool, then extract explicit commitments - who promised what to whom - with direction (inbound: promised to the owner; outbound: the owner promised), optional deadline, counterparty, and per-item confidence.

How it works 20 steps

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

Process (20 steps)

■ transcribed ?

skip: config.source is_empty

■ file_ref default

Apply default

■ has_file compare

Check if file ref has data

■ extract ?

skip: step.has_file not_equal true

■ defaults defaults

Apply defaults

■ overall_below_threshold compare

Check if extraction confidence is less than setting confidence threshold

■ commitment_rows_raw map

Extract from each item

■ commitment_rows_lc map

Extract direction from each item

■ commitment_rows_dir map

Extract direction from each item

■ commitment_rows_with_due map

Extract promised_for_at_iso from each item

■ commitment_rows_ready filter

Filter by condition

■ commitment_rows_clamped map

Extract confidence from each item

■ commitment_rows slice

Take a subset of

■ skip_writes any

Check if any of the conditions are true

■ create_commitments ?

skip: step.skip_writes equal true

■ commitment_count count

Count items in commitment rows

■ commitment_create_result_count count

Count items in create commitments

■ commitment_write_attempts ternary

Apply ternary

■ commitment_table to_table

Format results as a data table

■ card to_summary

Format results as a summary card

Settings 5 configurable

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

title setting =

source setting =

max items setting = 12

write mode setting = preview

confidence threshold setting = 0.7