memory-ingest

To turn meeting transcripts into structured knowledge, this Skill parses unstructured text, extracts key entities and action items, and creates linked source notes within Basic Memory to make information searchable and actionable.

Extracting entities from conversation logs involves identifying people, organizations, and topics, then searching for existing matches to propose new Basic Memory entities for approval to maintain knowledge graph integrity.

Yes, you can process pasted documents and unstructured text. The Skill understands various input formats, parsing them to create structured notes with observations, relations, and captured action items.

Yes, the entity extraction process searches for existing matches within Basic Memory first, then proposes new entities for approval before creating source notes to ensure knowledge graph integrity.

To link extracted action items to existing client notes, the Skill identifies attendees and decisions in the transcript, then automatically structures the meeting note and links it to existing Basic Memory entities.