Skill

Draft a reply to a routine inbound

Drafts a contextual response to one inbound message in the scholar's voice, surfacing decision-needed items rather than auto-sending.

In / outClassified inbound message + scholar's voice context → reviewed-ready draft reply

You might say…

I know roughly what I want to say — I just need a polished draft I can review and fire off.

What it does

Human-in-loop single LLM call that drafts a contextual response to one inbound message (decline, hold, acknowledge, forward-with-note) in the scholar's voice, surfacing items that need a decision rather than auto-sending. Used when working a message that warrants a human-reviewed reply.

Trigger: Use when working a message that warrants a human-reviewed reply rather than auto-send.

Recognise the problem?

The primitives are the commodity part. The fastest next step is a conversation about composing them into something that works for you.

Start a conversation