What problem does it solve? After a voice call ends, the transcript, audio, and key decisions are easily lost unless someone manually writes them up. This Skill turns every call into a structured brain page with a summary, verbatim transcript, audio link, and entity cross-links, then posts the summary to the operator's messaging surface. ## Core Features & Use Cases - Six-step post-call pipeline: captures audio via MediaRecorder, transcribes with Whisper, summarizes with an LLM, writes a brain page under meetings/ or voice-calls/, cross-links mentioned people and companies, and posts the summary to Telegram, Slack, or Discord. - Idempotent dual firing paths: supports both a persona-initiated mid-call log_to_brain tool call and a call-end handler, where the second invocation updates rather than duplicates the page. - Use Case: After ending a 2-minute voice call with the Venus persona, the Skill produces meetings/2026-05-17-call-venus.md containing frontmatter, a 3-5 sentence summary, the verbatim block-quoted transcript, the audio file link, and timeline entries on each mentioned person's page. ## Quick Start After the call ends, ask the agent to summarize the call and file the transcript into the brain with links to everyone mentioned.
