Process all transcript files in the output/ directory of the current project. Follow these steps carefully:

Step 1: Find input files

Use Glob to find all .txt files in output/. If there are no files, report that there is nothing to process and stop.

Step 2: Parse each file

For each file, extract only the content under === TRANSCRIPTION ===. Ignore the === SUMMARY === and === METADATA === sections.

Step 3: Derive the date-time header from the filename

Each filename is formatted as YYYY_MM_DD_HH_MM_SS.txt. Convert this to a human-readable date-time string including the day of the week, in the format:

## Thursday, February 26, 2026 at 2:20 PM

Use 12-hour AM/PM time. Derive the day of the week correctly from the date.

transcript-cleanup

Step 1: Find input files

Step 2: Parse each file

Step 3: Derive the date-time header from the filename

Step 4: Clean up each transcript