swarm hub / onboarding / sessions

Your 6 Input Sessions

Schedule them. Prepare for them. Find your recordings. Turn them into features.

Why these six 30-minute Zooms matter more than anything else this month

You will never get this time back. John, Mark, and Ewing each give you 30 minutes. That is 90 minutes of expert input per intern. Used well, it spawns a quarter of features. Used poorly, you walk away with notes you can't act on. Your packages (Charlie / Bear) have the question banks. This page is about the operational side: scheduling, recording, and turning the recording into a feature spec.

Step 1. Schedule all six (today)

Don't ask permission. Pick three slots each. They've blocked time.

Charlie : John

The calling discipline. Years of cold open mastery.

Charlie : Mark

The operational reality. What slows him down daily.

Charlie : Ewing

The vision. Where the meetings page goes next.

Bear : John

Pipeline discipline across years of running deals.

Bear : Mark

Daily deal touching. What busywork to kill first.

Bear : Ewing

The 5x deal coverage vision. What you're actually building toward.

Open Google Calendar Email Ewing

Step 2. Show up prepared

Read the prep pack the night before. Re-read it 30 minutes before the call.

The session will be recorded by Fireflies automatically. John, Mark, and Ewing all have Fireflies on their accounts. The bot joins the Zoom. You don't have to do anything. The transcript will appear at os.chapter.guide/meetings within an hour after the call ends.

Because Fireflies sync is automatic, you should never physically open the Fireflies website to retrieve a transcript. Everything you need lands in our system. The Meetings page is your interface to your own recordings.

Step 3. Process each session into a requirements artifact

Within 24 hours of each call. The output of this step is what you build from.

1

Find your recording

Open os.chapter.guide/meetings. Filter by participant (your name + the expert's name). Click your session. The transcript, summary, and decisions are already there. Fireflies populated them automatically.
2

Run extract

On the meeting detail page, click Extract. The system runs Claude on the transcript and pulls out: themes, pain points, feature ideas, surprises, decisions, and follow-ups. Output lands in transcript_extractions and renders right on the page. Cost per extraction: about $0.06.
3

Fill the writeup template

Use the template below. Don't skip the "one thing they want but didn't say" line. That's where the best features come from.
4

Tag it for the research output page

Set the meeting's classification to internal_training and add the tag intern_input_session. The research output page picks it up automatically and aggregates findings across all six sessions.
5

Surface the feature ideas

Each session produces 2 to 5 feature ideas. Don't try to build them all. Pick the one that solves the loudest pain point AND fits in your first 2 weeks. Create a task at /review/tasks with the title and a 3-sentence description. Ewing reviews. If approved, that becomes your next build.

Writeup template

Fill this in within 24 hours. Save in your package's notebook section.

Session: <Charlie or Bear> : <John, Mark, or Ewing> Date: YYYY-MM-DD Goal stated: <the one sentence you opened with> Fireflies link: os.chapter.guide/meetings/<id> Three themes that came up: 1. ... 2. ... 3. ... One thing that surprised me: One thing they want but didn't say (my interpretation): Three feature ideas that came out of this: 1. <name> — <one sentence pain point> — <one sentence what it does> 2. 3. What I'm doing in the next 7 days based on this: Open question I want to ask in the follow-up:

The research output page

Where the findings from all six sessions live and compound.

Once you tag a meeting as intern_input_session, it appears on the research output page at os.chapter.guide/intern-research. That page aggregates:

The page is the canonical source for what you build next. Mark, John, and Ewing read it. They'll comment. They'll add things they remember from the call that you missed. Treat it as a living document, not a one-time output.

Anti-patterns we'll catch you doing

Now go open your package

Charlie's package has your prep packs and your first AI build. Bear's package has yours.