The first comprehensive event management skill collection for Claude and compatible AI agents.
12 AI agent skills (10 specialist + 2 meta) covering the complete B2B event lifecycle — from initial concept through post-event analytics. Each skill is a self-contained module with structured prompts, reference data, and quality gates. Together they handle conferences, trade shows, product launches, sales kickoffs, developer conferences, industry summits, roadshows, VIP events, webinars, and community meetups.
Built for event managers, marketing teams, and operations professionals who plan and execute B2B events. Event management has been underserved in the AI skills ecosystem. This collection changes that.
Organized by the four phases of an event lifecycle. Every skill is self-contained — pick the ones you need.
| Skill | What it does |
|---|---|
| event-planner | Strategic planning, timelines, checklists, RACI matrices, risk registers |
| event-budget | Budget breakdowns, calibration-based benchmarks, scenario modeling, ROI forecasts |
| event-speakers | CFP creation, speaker outreach, agenda building, diversity framework |
| event-sponsors | Sponsorship packages, prospectus documents, exhibitor guides, fulfillment |
| Skill | What it does |
|---|---|
| event-marketing | Promotion strategy, channel playbooks, registration funnels, campaign briefs |
| event-content | Emails, social posts, website copy, MC scripts, anti-AI writing standards |
| event-registration | Registration forms, ticket tiers, check-in operations, GDPR compliance |
| Skill | What it does |
|---|---|
| event-logistics | Venue evaluation, vendor RFPs, run-of-show, emergency procedures |
| event-exhibition | Trade show booth planning, sponsorship evaluation, lead capture, post-show ROI |
| Skill | What it does |
|---|---|
| event-analytics | KPI frameworks, post-event surveys, ROI calculations, lessons learned |
| Skill | What it does |
|---|---|
| event-orchestrator | Multi-skill workflow coordination, conditional routing, failure recovery |
| event-reviewer | Cross-skill consistency checks, prescriptive remediation |
Standalone, not monolithic. Every skill works on its own — event-content doesn't need event-planner. Take one, take all 12.
Anti-AI writing standard. Content-generating skills pass through a writing-quality gate before output: no uniform sentence lengths, no formulaic transitions, no symmetrical bullet lists, no hype tone. Copy you can actually send.
Progressive disclosure. A skill activation costs roughly 500–2,000 tokens for the workflow, plus only the reference data the specific request needs. No oversized prompts.
Verification gates, not vibes. Every skill has a Step 3.5 quality check before presenting output. Budgets balance, timelines are dependency-correct, risk registers cover the right categories.
Real test scenarios. The skills were stress-tested across a range of B2B event formats — flagship conferences, partner co-marketing, lead-gen webinars, trade-show booths, executive dinners, multi-country roadshows, and C-level roundtables.
git clone https://github.com/simonkoehler/event-skills.gitThen either:
- Use
--add-dir /path/to/event-skillswhen launching Claude Code - Or copy the skill folders into your project's
.claude/skills/directory
Copy the skill folders you need into ~/.claude/skills/:
cp -r event-skills/event-planner ~/.claude/skills/
cp -r event-skills/event-content ~/.claude/skills/
# ... repeat for each skill you wantEach skill is fully standalone. Copy just the folder you need — no dependencies on other skills required.
Once installed, just ask naturally:
"Plan a 300-person tech conference in Munich for October"
"Create sponsorship packages for our annual summit"
"Write the email sequence for our developer conference registration"
The right skill activates based on your request. For complex events, multiple skills work together — start with event-planner for the strategic foundation, then use specialized skills for execution details.
This collection uses progressive disclosure to stay token-efficient:
- Level 1 — Name + Description (~100 tokens): Always loaded. The YAML frontmatter tells the agent when to activate the skill.
- Level 2 — SKILL.md body (<500 lines): Loaded when the skill triggers. Contains the full workflow, output templates, verification gates, and examples.
- Level 3 — Reference files (on-demand): Loaded only when a specific output needs them. Detailed templates, benchmarks, checklists, and frameworks that would waste context if loaded upfront.
This means a single skill activation costs roughly 500-2,000 tokens for the workflow, plus only the reference data actually needed for the user's request.
These skills follow the Agent Skills Open Standard and work with:
- Claude Code and Claude Cowork
- OpenAI Codex CLI
- Cursor
- Gemini CLI
- Any tool that supports the
SKILL.mdformat
See COLLECTION.md for a detailed map of how the 12 skills connect, dependency relationships, and routing guidance by event type.
PRs welcome. Every skill must pass this quality checklist before merging:
- SKILL.md under 500 lines
- YAML description under 1,024 characters
- All referenced files exist locally in the skill directory
- No cross-skill file references
- Verification gate with skill-specific checks
- 3+ examples with realistic numbers
- 5+ edge cases with concrete handling instructions