Read-only AssemblyAI transcript review workflow. It retrieves one transcript by ID, tables chapters, highlights, sentiment, entities, and utterances when present, and summarizes readiness and insights. It never creates transcripts, uploads files, deletes transcripts, registers webhooks, uses streaming, or calls arbitrary AssemblyAI routes.

How it works 16 steps

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

Process (16 steps)

■ normalized_transcript_id trim

Trim whitespace from setting transcript id

■ transcript ?

■ chapters default

Apply default

■ highlights_raw default

Apply default

■ sentiment_raw default

Apply default

■ entities_raw default

Apply default

■ utterances_raw default

Apply default

■ highlights slice

Take a subset of

■ sentiment slice

Take a subset of

■ entities slice

Take a subset of

■ utterances slice

Take a subset of

■ summarize ?

■ chapter_table to_table

Format results as a data table

■ highlight_table to_table

Format results as a data table

■ utterance_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.

focus setting = Summarize transcript status, readiness, summary text, chapters, highlights, sentiment, entities, speakers, action-worthy moments, and follow-up questions. If the transcript is not completed, explain what is still pending from the AssemblyAI response.

row limit setting = 20

assemblyai setting =

max length setting = 900

transcript id setting =