meeting

To summarize meeting transcripts automatically, place the raw transcript files into Raw-Materials/Meeting-Transcripts/. The system processes the oldest queued file, assesses duration and participant count, and outputs a structured narrative capturing decisions and action items to Curated-Context/Meeting-Insights/.

Extracting action items and decisions from meeting minutes is handled by selecting minimal, standard, or full templates based on substance. The process scans recent context files for cross-references, archives the original transcript, and writes a balanced summary to prevent redundancy.

Processing multiple meeting transcripts simultaneously is not supported. The system strictly targets the oldest file in the queue directory to ensure sequential processing, assesses its metadata, writes the structured summary, and archives the original before moving to the next.

Meeting summarization supports cross-referencing by scanning recent files in Curated-Context/Meeting-Insights/ before writing. This ensures new summaries incorporate confirmed facts, update entity stubs, and connect current decisions to existing knowledge without redundancy.

A full template for meeting summaries is needed when the transcript contains high substance requiring detailed context. The system automatically selects minimal, standard, or full templates by assessing duration, participant count, and content depth before generating the structured narrative.

Processed meeting transcripts are archived after summarization to prevent reprocessing. The resulting structured summary is stored in Curated-Context/Meeting-Insights/YYYYMM/ organized by year and month, feeding into morning journals and memory updates when decisions emerge.