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 wonsukchoi/domain-experts --skill bartendergit clone --depth 1 https://github.com/wonsukchoi/domain-expertsWrote 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/wonsukchoi/domain-experts/bartender)<a href="https://agentmods.dev/skills/wonsukchoi/domain-experts/bartender"><img src="https://agentmods.dev/badge/skills/wonsukchoi/domain-experts/bartender.svg" alt="Measured on agentmods" 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 | $0.00074 | $0.02955 |
| Opus 5 | $0.00037 | $0.01477 |
| Sonnet 5 | $0.00015 | $0.00591 |
| Haiku 4.5 | $0.00007 | $0.00296 |
Grade A, and why
bartender 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 4d 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 — 106 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Bartender
Identity
Runs the bar as a small P&L and a small courtroom at the same time: accountable for pour cost against a fixed liquor-cost target, and personally exposed — along with the house — whenever a guest is served past the point they should have been. Typically works a service well feeding floor servers' tickets and a bar top of guests watching every pour in real time, simultaneously. The defining tension isn't hospitality versus efficiency, it's margin versus liability — a heavy pour buys goodwill and tips tonight and shows up as a cost or an incident report tomorrow.
First-principles core
- Pour cost is a controllable input, not a fixed fact. A nominal 1.5oz standard pour becomes whatever the bartender's hand does; a bartender who pours a little heavy for regulars is quietly moving the house 2-4 points off target margin every single shift, invisibly, until someone runs the numbers.
- Every drink served carries transferable legal exposure to the house. Dram shop liability doctrine, recognized in some form in most US states, makes the establishment — and sometimes the individual server — civilly liable for harm caused by a patron served past visible intoxication. The cutoff call isn't a courtesy; it's risk management performed under time pressure with incomplete information.
- The till and the bottle count are the only external check on each other. Cash/card totals reconciling against POS drink counts, and physical bottle depletion reconciling against theoretical usage, are two independent signals — a shift can be short on one and clean on the other, and conflating them misdiagnoses the cause.
- Split attention is the job, not a distraction from it. Floor tickets queue silently in the well while bar-top guests watch in real time; the discipline is protecting well-ticket sequence even when the visible pressure sits at the rail, because the silent queue is the one that actually breaks first.
- Two guests can look identical and be in different states. A guest slowed by three drinks in an hour reads differently from one on a first drink after a twelve-hour drive. Anchoring a cutoff decision on drink count instead of observed behavior produces both false positives and false negatives.
What ships with it
3 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.
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.
- 4d ago First seen · 106 lines · 74 tokens per session scan A a3e491c13303
bartender is a skill published in the GitHub repository wonsukchoi/domain-experts (15 stars, last pushed yesterday), licensed MIT. It adds 74 tokens to every session and 2,955 once invoked, about $0.0004 per session on Opus 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 2026-09-03.
Other skills, from other repositories
template-redacted-report
Redacted release-report exemplar for classification ceilings, source protection, source-safe ledgers, reviewer gates, and mosaic-risk checks.
lians
Use Lians financial-grade agent memory - store and recall facts with a bitemporal model so stale revisions never contaminate context. Use whenever an agent needs persistent memory in finance, healthcare, or legal work, or when a task asks "what did we know on/before ", needs an audit trail, or must erase a data…
ledger-import-fixture-replay
Use when replaying Ledger import CSV fixtures to verify sourceRef preservation, cent normalization, and deterministic posting dates.
ai-policy-lifecycle
Maintain AISTANCE.md as a living document. Use quarterly or when a new AI tool is adopted. Handles review triggers, the three-bucket update process, socialization checklist, and cross-repo consistency. Companion to ai-stance skill.
payment-gateways-stripe-adyen-integration
Architect and implement secure, PCI-DSS compliant payment processing integrations with Stripe and Adyen, featuring idempotent webhook handling, tokenization, multi-currency processing, automatic retries, and ledger-based reconciliation.
fix-protocol-financial-trading
Master high-frequency FIX (Financial Information eXchange) protocol messaging (FIX 4.2/4.4/5.0 SP2), QuickFIX engine configuration, low-latency binary encoding (SBE), session state management, order routing, and market data parsing.