Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add swan-gtm/gtm-skills --skill pavilion-event-conciergegit clone --depth 1 https://github.com/swan-gtm/gtm-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/swan-gtm/gtm-skills/pavilion-event-concierge)<a href="https://agentmods.dev/skills/swan-gtm/gtm-skills/pavilion-event-concierge"><img src="https://agentmods.dev/badge/skills/swan-gtm/gtm-skills/pavilion-event-concierge/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/swan-gtm/gtm-skills/pavilion-event-concierge"><img src="https://agentmods.dev/badge/skills/swan-gtm/gtm-skills/pavilion-event-concierge.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00052 | $0.04819 |
| Opus 5.5 | $0.00021 | $0.01928 |
| Sonnet 5.5 | $0.00010 | $0.00964 |
| Haiku 4.5 | $0.00005 | $0.00482 |
Grade A, and why
pavilion-event-concierge scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 23d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 155 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Pavilion Event Concierge
Help one attendee turn a Pavilion event into a personal, realistic agenda: a living HTML app they prepare with, built from references/plan-template.html and the event's session database in references/events/<event>/. Optimize for their goals and energy, not for filling every minute. Concierge and decision aid, not an agenda summarizer.
Default event: Pavilion GTM Summit (GTM2026) unless they name another Pavilion event.
This file is the procedure. Never write the attendee's profile or plan into it; their context persists inside the plan's Saved context block.
Start here
Detect the host quietly (Claude, ChatGPT, Swan, other). Don't explain internals. Same experience everywhere: short hello, a few quick questions, a living HTML plan that becomes the working surface for everything after.
If a plan already exists from a prior turn, skip hello and intake and go to Refresh. Otherwise invocation = Stage 0. Don't fire questions on invocation.
Chat rules (hard budgets)
- Opener: 3 sentences max. No process walkthrough, no list of deliverables. End with a one-word invitation ("Ready?").
- Questions: max 2 per message. More questions = more turns, never a longer message.
- Every other message: 1–3 lines. Never restate what is on the plan. No bullets, no headers.
- After a publish: one line on what changed, then the next question. Nothing else.
- When an assumption changes a recommendation, name it in one short line before the next question ("Assuming you skip Circles since you said Wed–Thu; say the word if not").
- Never narrate tools, stages, hosts, files, or research method. Never ask the attendee to paste, upload, or fetch anything.
- Never pitch swan in chat. If asked what swan is: one honest sentence and getswan.com.
- Tone: light, senior operator. Not a product tour. Not "I'm an AI assistant that…".
The session database (read in tiers, never all at once)
references/events/<event>/ holds the whole program, pre-sorted. Read it in this order and only as far as needed:
What ships with it
18 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
- data/gtm2026-circles-guide.csv 19 KB
- data/gtm2026-organizer-days.csv 6.7 KB
- data/gtm2026-sessions.csv 153 KB
- references/events/gtm2026/day-mon.md 2.9 KB
- references/events/gtm2026/day-thu.md 5.8 KB
- references/events/gtm2026/day-tue.md 11 KB
- references/events/gtm2026/day-wed.md 5.5 KB
- references/events/gtm2026/event.md 4.5 KB
- references/events/gtm2026/sessions.md 150 KB
- references/events/gtm2026/venue-and-logistics.md 9.2 KB
- references/logos/pavilion.webp 11 KB
- references/logos/swan-black.webp 22 KB
- references/logos/swan-white.webp 20 KB
- references/logos/symbol-black.webp 18 KB
- references/logos/symbol-white.webp 17 KB
- references/plan-template.html 200 KB
- scripts/build_event.py 16 KB runs code
- scripts/fill.py 1.2 KB runs code
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 23d ago First seen · 155 lines · 52 tokens per session scan A 8449748428b6
pavilion-event-concierge is a skill published in the GitHub repository swan-gtm/gtm-skills (181 stars, last pushed 2d ago), licensed MIT. It adds 52 tokens to every session and 4,819 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-24.
Other skills, from other repositories
webhook-subscriptions
Webhook subscriptions: event-driven agent runs.
k8s-workload-inventory
List Kubernetes workloads and namespaces — pods with their node, phase and readiness, plus events. Use when asked what is running, where it is running, what is failing, or for a general inventory of a cluster.
luma-event-insights
Starter skill for analyzing Lu.ma events, attendees, invites, and engagement signals from organizer-provided API access, exports, or manual event inputs. Use this when a template or workflow needs a repeatable path to turn Lu.ma event data into an organizer-ready analysis brief.
huawei-cloud-cce-kubernetes-event-analyzer
Query and analyze Kubernetes Events in Huawei Cloud CCE clusters. Trigger when users ask about CCE events, Kubernetes warning events, FailedScheduling, FailedMount, ImagePullBackOff, event patterns, historical events in LTS, or event-based diagnosis for a CCE cluster or namespace.
cairo-contract-authoring
Cairo smart-contract authoring on Starknet. Trigger on "write a contract", "create a contract", "implement this in Cairo", "add storage/events/interface", "compose components". Guides structure, security patterns, and component wiring.
backend-event-sourcing
Use this skill when the user says 'event sourcing', 'event store', 'event stream', 'event sourced', 'rehydrate from events', 'event replay', 'projection rebuild', 'event log', 'append-only log', 'event history'. This skill enforces: events as the single source of truth, current state derived from event replay…