From cowork-research
Pull a Fathom transcript and extract decisions, action items, open questions, and notable quotes. Reads about-me/business-brain.md (for ICP context — what the user cares about). Uses Fathom MCP. Lands in reference/research/meetings/<date>-<title>.md. Optionally drafts to chosen destination.
How this skill is triggered — by the user, by Claude, or both
Slash command
/cowork-research:meeting-recapThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Pull a Fathom meeting transcript and extract decisions + action items + open questions.
Pull a Fathom meeting transcript and extract decisions + action items + open questions.
about-me/about-me.mdabout-me/business-brain.md (for ICP context — what action items matter to THIS user)templates/meeting-recap-template.mdprojects/research/research-preferences.md (recap depth: executive / detailed / verbatim)mcp__claude_ai_Fathom__get_recording_by_url; call_id → get_recording_by_call_id; recording_id → get_meeting_summary directlyget_meeting_transcript with the resolved recording_id (for timestamped quotes) AND get_meeting_summary (for AI-generated summary as a starting point)templates/meeting-recap-template.md shapereference/research/_index.mdprojects/research/memory.mdreference/research/meetings/<date>-<title>.md (the recap)get_meeting_transcript first (per Fathom MCP grounding rule)Tactical, scannable. Decisions and actions are the load-bearing parts — the rest is supporting. Quotes only when they capture something the summary can't.
npx claudepluginhub automatedmarketer/cowork-research --plugin cowork-researchCreates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.