@dklineii - Dk Lineii Morning Brief

Open original

@dklineii on X This is a skills file for building an AI-generated morning brief that reads your calendar, email, tasks, team chat, and metrics, then returns a one-screen plan. It walks through six reads (Day, Decisions, People, Meetings, World, Move) and six prompts to run it, including a "people radar" for spotting team members who need attention.

  • The brief runs as a 15-minute daily routine, ideally before opening email, so your priorities set the day rather than your inbox.
  • Four rules govern output: name actual meetings and times instead of generic advice, flag missing data instead of stopping, actually research external contacts, and never fabricate.
  • The People read is presented as the key differentiator for managers, since it covers team check-ins, commitments from 1:1s, and early signs of disengagement.
  • The author suggests pasting the article into Claude Cowork to build the skill, connecting your tools, and scheduling it to run overnight.

The piece is a how-to for building an AI-generated "morning brief," pitched at managers who start their days buried in email. The core claim is that you should spend your first fifteen minutes reviewing a structured plan before touching your inbox. The brief the author describes has six sections: the Day (calendar and tasks merged, with open blocks marked), the Decisions (what's blocked on you, ordered by cost of delay), the People (who needs you, who committed to what in the last 1:1, who has gone quiet), the Meetings (one line each on purpose, desired outcome, and a question to ask), the World (three to five outside developments that touch your work), and the Move (one high-leverage change to the day, with a drafted message to make it happen). The author lays out four rules for keeping such a brief trustworthy: make concrete recommendations naming real meetings and times, flag missing inputs instead of stopping, actually research rather than summarize titles, and admit when something is unknown rather than guessing. Without these, the author argues, a brief reads with false confidence and is worse than none. The implementation is a prompt you paste into an LLM tool like Claude Cowork, connect to your calendar, email, chat, tasks, and metrics, and schedule to run overnight. The article includes six prompts: the main brief builder, a "people radar" that picks the one team member most needing attention, prompts for meeting prep and world-news filtering, and two for grading the day and closing loose ends from yesterday.

Ask about this article