Install any skill in seconds. Free to start, no credit card required.
Get Started Free →When the user wants to host a tech event, meetup, workshop, or community gathering — especially using Luma. Also use when the user mentions "host an event", "meetup", "workshop", "Luma event", or "community event".
.claude/skills/mkurman-event-hosting/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-07 | ✗→✓ | ▲ Improved | 73% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 0% | 0% |
| case-14 | ✗→✓ | ▲ Improved | 49% | 0% |
| case-19 | ✗→✓ | ▲ Improved | 35% | 0% |
| case-03 | ✓→✓ | = Same ✓ | 53% | 0% |
|---------|-------|-------| | 6:00 PM | Doors open | — | Networking, food/drinks | | 6:15 PM | Welcome | Founder] | Context, housekeeping | | 6:20 PM | Talk 1 | Speaker] | Topic] | | ... | ... | ... | ... |
| Channel | Action | Timeline | |---------|--------|----------| | Luma | Publish event page | 3 weeks before | | LinkedIn | Founder post + speaker reshares | 2 weeks before | | Communities | Post in specific communities] | 2 weeks before | | Direct outreach | Personal invite to 20 anchor attendees | 2 weeks before | | Reminder | Luma auto-reminder + personal note | 1 day before |
## Frameworks & Best Practices
**Event formats that work for startups:**
| Format | Best For | Typical Size | Effort |
|--------|----------|-------------|--------|
| Fireside chat | Thought leadership, intimate discussion | 20-50 | Low |
| Panel | Diverse perspectives, networking draw | 30-100 | Medium |
| Workshop | Teaching, product education, lead gen | 15-30 | Medium |
| Demo night | Showcasing products, community building | 30-80 | Medium |
| Hackathon | Developer community, product feedback | 20-100 | High |
| Dinner / small gathering | Investor/executive networking | 8-15 | Low (but high cost) |
**Luma-specific tips:**
- Use the "Require Approval" feature to curate attendees (quality > quantity)
- Enable "Ask a Question" during registration to qualify attendees
- Set up co-hosts so speakers can also invite their networks
- Use Luma's calendar subscription feature for recurring events
- Luma's built-in reminder emails have high open rates — don't duplicate with external email
**The 40% rule:** Expect 40-60% of RSVPs to actually attend for free events. For paid events, expect 80-90%. Overbook accordingly.
**Making events a growth channel:**
- Host monthly, not "whenever" — consistency builds an audience that returns
- Each event should have a natural next step (join our community, try the product, attend the next event)
- Record everything — one event becomes 5-10 pieces of content
- The best events are conversations, not presentations — leave 30%+ of time unstructured
**Common mistakes:**
- Making it about your product instead of your audience's interests
- Inviting everyone instead of curating for quality
- No follow-up after the event (the event is the beginning, not the end)
- Over-programming — leave room for serendipity and networking
- Not starting the series until everything is "perfect" — first event can be 15 people in a coffee shop
## Related Skills
- `community-discovery` — find the right communities to promote your event
- `founder-thought-leadership` — events are a powerful thought leadership channel
- `content-strategy` — repurpose event content into blog posts, social, and newsletters
- `employer-brand` — hosting events signals culture and attracts talent
## Examples
**Prompt:** "Help me create a Luma event for a monthly AI founders meetup in SF."
**Good output includes:** Event page copy with a compelling title and description, run of show for a 2-hour evening event, promotion plan targeting AI-focused communities, and a follow-up template.
**Prompt:** "I want to host a virtual workshop teaching developers how to use our API."
**Good output includes:** Workshop structure with live coding segments, Luma event page optimized for developer audience, promotion through developer communities (from community-discovery), and a lead capture strategy.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 15,977 | 11,628 | -27% | 1 | 1 | 0% | 2,520 | 2,715 | +8% | 0 | 0 | — |
case-02 | fail→fail | 22,160 | 18,525 | -16% | 1 | 1 | 0% | 3,426 | 3,792 | +11% | 0 | 0 | — |
case-03 | pass→pass | 17,879 | 22,398 | +25% | 1 | 1 | 0% | 2,545 | 3,884 | +53% | 0 | 0 | — |
case-04 | pass→pass | 23,385 | 20,020 | -14% | 1 | 1 | 0% | 3,764 | 4,059 | +8% | 0 | 0 | — |
case-05 | pass→pass | 17,955 | 19,415 | +8% | 1 | 1 | 0% | 2,643 | 3,771 | +43% | 0 | 0 | — |
case-06 | pass→pass | 9,602 | 11,356 | +18% | 1 | 1 | 0% | 1,610 | 2,750 | +71% | 0 | 0 | — |
case-07 | fail→pass | 8,348 | 9,072 | +9% | 1 | 1 | 0% | 1,310 | 2,265 | +73% | 0 | 0 | — |
case-08 | fail→pass | 12,738 | 6,570 | -48% | 1 | 1 | 0% | 1,991 | 1,996 | +0% | 0 | 0 | — |
case-09 | pass→pass | 16,949 | 16,801 | -1% | 1 | 1 | 0% | 2,470 | 3,495 | +41% | 0 | 0 | — |
case-10 | fail→fail | 15,841 | 17,841 | +13% | 1 | 1 | 0% | 2,498 | 3,690 | +48% | 0 | 0 | — |
case-11 | pass→pass | 12,861 | 14,248 | +11% | 1 | 1 | 0% | 1,908 | 3,104 | +63% | 0 | 0 | — |
case-12 | pass→pass | 13,509 | 13,476 | -0% | 1 | 1 | 0% | 1,987 | 3,014 | +52% | 0 | 0 | — |
case-13 | pass→pass | 11,678 | 11,879 | +2% | 1 | 1 | 0% | 1,755 | 2,755 | +57% | 0 | 0 | — |
case-14 | fail→pass | 11,757 | 10,762 | -8% | 1 | 1 | 0% | 1,792 | 2,668 | +49% | 0 | 0 | — |
case-15 | fail→fail | 12,178 | 12,669 | +4% | 1 | 1 | 0% | 1,882 | 2,828 | +50% | 0 | 0 | — |
case-16 | pass→pass | 11,048 | 12,173 | +10% | 1 | 1 | 0% | 1,699 | 2,681 | +58% | 0 | 0 | — |
case-17 | pass→pass | 15,967 | 13,173 | -17% | 1 | 1 | 0% | 2,639 | 3,033 | +15% | 0 | 0 | — |
case-18 | fail→fail | 11,222 | 8,278 | -26% | 1 | 1 | 0% | 1,680 | 2,066 | +23% | 0 | 0 | — |
case-19 | fail→pass | 12,110 | 9,603 | -21% | 1 | 1 | 0% | 1,767 | 2,384 | +35% | 0 | 0 | — |
case-20 | pass→pass | 16,335 | 16,961 | +4% | 1 | 1 | 0% | 2,416 | 3,387 | +40% | 0 | 0 | — |
case-21 | pass→pass | 13,134 | 10,947 | -17% | 1 | 1 | 0% | 2,021 | 2,575 | +27% | 0 | 0 | — |
case-22 | pass→pass | 9,354 | 9,229 | -1% | 1 | 1 | 0% | 1,508 | 2,392 | +59% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +18 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.