agentleFS
guerrilla2799guerrilla2799/event-and-scale-os/skills/meeting-engine/SKILL.md

meeting-engine

Skill2 starsChanged 3 months agoMITView on GitHub

Use for any conference or multi-day event, or when the user says "book meetings at the event", "we need meetings scheduled", "nobody booked anything", "matchmaking platform", "who are we meeting at the show". Builds the pre-booked meeting operation: the target derived from real attendance, outreach sequenced around the matchmaking window, and per-person schedules. Writes workspace/events/<event>/meetings.csv.

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 3 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    No deck attached, no calendar link in the first message for senior recipients, no product description.

    line 70
  2. Distinctive:

    Every meeting gets a note the same day: what was discussed, the next step, and the date.

    line 90
  3. Distinctive:

    Same-day notes: No exceptions

    line 102
  4. Distinctive:

    No fabricated personalization: Do not claim to have read something you have not

    line 103

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Meeting Engine

Pre-booked meetings are the primary metric for any event where money changed hands. This is the operation that produces them.

When to use

  • Any sponsored conference, immediately after signature
  • Any conference you are attending without sponsoring
  • An owned event with a VIP track that needs scheduled conversations
  • Post-mortem, when an event produced almost no meetings

Inputs

  • Reads: workspace/events/<event>/contract.md (the matchmaking answers), workspace/audience/target-list.csv, workspace/program/charter.md
  • Needs from user: who is attending from your side and for how long, the matchmaking platform details, and the attendee or registration list if the organizer shares one

Workflow

1. Derive the target, do not pick it

Expected attendance (confirmed with a human at the organizer)
  - sponsor and exhibitor staff
  = buyer-side population
  x share in ICP industries
  x share at or above the seniority bar
  = senior ICP pool

Target meetings / senior ICP pool = penetration rate

Order matters. Net the staff out first. They are entirely non-prospects, not a slice distributed across ICP and seniority like real attendees, so subtracting them at the end removes them twice. On 1,000 attendees, 30% ICP, 50% senior, 100 sponsor staff: the wrong order returns 50, the right order returns 135.

25 to 35% penetration is aggressive and achievable with a stage slot and a long matchmaking window. Above 50% is fantasy. Below 10% means the event is wrong or nobody is trying.

Then check it against staff capacity. Two people running back-to-back thirty-minute meetings across a seven-hour day, minus sessions they must attend and a lunch, is a hard ceiling. If the target exceeds capacity, either bring more people or lower the target. Do not book meetings nobody can hold.

Validate weekly. Many organizers share a running registration list. Check the target against real registered names each week rather than defending an abstract number.

2. Sequence outreach around the matchmaking window

The date the platform opens decides everything. Get it from a human, not from the booklet.

WhenWhat
T-6, before the platform opensWave 1 to your own named list, by email, from account owners. "We will both be there, can we find thirty minutes"
Platform open dayWave 2 inside the platform, to everyone matching ICP. Be early. The first week is when calendars are empty
T-3Wave 3, second touch on non-responders, both channels
T-1Final push, plus confirmation of everything already booked
On siteSame-day booking for people you meet. Keep slots open for this

Landing wave 1 before the platform opens is the correct sequence. Those people arrive at the platform already knowing your name.

Keep 20 to 30% of slots unbooked. The best conversations at a conference are the ones you did not plan, and a fully booked calendar means you cannot take them.

3. Write the outreach

Short, specific, and about them being there rather than about your product.

Subject: <Event name>, Thursday

<Name>, saw you're speaking at / attending <event> on <date>.
<One clause on why them specifically: their session, a recent announcement, a
shared connection, something at their company.>
We're on stage at <time> talking about <topic>. Any chance of 30 minutes
either side of that? Happy to work around your schedule.

No deck attached, no calendar link in the first message for senior recipients, no product description. Every invitation runs through writing-quality.

4. Use the transferable passes as the strongest outreach you have

A comped pass is a reason to call a prospect that has nothing to do with your product, and it works on people who ignore everything else. Assign every pass to a named person, chase the registration personally, and set an internal cutoff a day before the real one. Passes unredeemed at the deadline are the most expensive waste in an event budget.

5. Build the schedule

Per person, per day, one page per meeting:

  • Time, place, and how to find each other. "By the coffee station" is not a location at a conference
  • Who is attending from both sides
  • One page on the account: why they matter, history, the point of view, the ask
  • The next step you want, written before you walk in

Distribute the night before, in the huddle. Never on the morning of.

Location matters more than people expect. A booth is a bad meeting location. A quiet corner, a hotel lobby nearby, or a booked meeting room is better. If the venue offers meeting rooms, book them the day they open.

6. Run confirmations

A meeting booked three weeks out and never mentioned again is a 50% show rate. Confirm at T-3 days and again the morning of, with the location and a phone number.

7. Capture on the day

Every meeting gets a note the same day: what was discussed, the next step, and the date. Notes written three days later are worth a fraction as much, and the follow-up written from them shows it.

Output

  • Writes: workspace/events/<event>/meetings.csv from templates/meeting-tracker-schema.csv, and per-person schedules
  • Prints: the derived target with its arithmetic, capacity check, the outreach calendar against the platform window, pass assignments, and current booked against target

Rules & quality bar

  • Derive the target from attendance arithmetic, then check it against staff capacity
  • Wave 1 lands before the matchmaking platform opens
  • Keep 20 to 30% of slots open for what you did not plan
  • Every transferable pass is assigned to a named person and chased personally
  • One page per meeting, distributed the night before
  • Same-day notes. No exceptions
  • No fabricated personalization. Do not claim to have read something you have not

Related skills

  • Reads: sponsorship-negotiation, audience-machine
  • Pairs with: conference-activation, staffing-and-briefing
  • Feeds: event-followup, event-measurement

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.