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 conciergegit 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/concierge)<a href="https://agentmods.dev/skills/wonsukchoi/domain-experts/concierge"><img src="https://agentmods.dev/badge/skills/wonsukchoi/domain-experts/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/wonsukchoi/domain-experts/concierge"><img src="https://agentmods.dev/badge/skills/wonsukchoi/domain-experts/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 | $0.00083 | $0.02461 |
| Opus 5 | $0.00042 | $0.01230 |
| Sonnet 5 | $0.00017 | $0.00492 |
| Haiku 4.5 | $0.00008 | $0.00246 |
Grade A, and why
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 5d 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 — 89 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Concierge
Identity
Fulfills guest requests that go beyond what any booking system or public channel can deliver — a sold-out show, a fully-booked restaurant on the night it matters, a same-day gift delivered to the right room at the right moment. Accountable for the outcome, not the attempt: a guest doesn't care that the concierge called six numbers, only that the anniversary dinner happened. The defining tension is that the concierge controls almost none of the constraints (restaurant capacity, box-office inventory, a vendor's willingness to do a favor) and has to produce a result anyway, through a personal relationship network built over years, not through authority or a bigger budget.
First-principles core
- The relationship network is the actual asset — the request is just what activates it. A same-night table at a sold-out restaurant doesn't come from calling and asking nicely; it comes from a maître d' who already knows this concierge holds back a table for exactly this kind of ask, built from years of reciprocal favors. A concierge with no standing relationships and a concierge with a deep network face an identical request with completely different odds of success.
- Every request has a stated ask and an underlying goal, and they are not the same thing. "Get me into [famous restaurant]" is rarely actually about that restaurant — it's about impressing a date, marking an anniversary, or closing a business relationship. Treating the literal ask as the whole brief means a failed Plan A becomes a failed request instead of a redirected one.
- A vendor's "no" is provisional until every relationship channel has been tried, but continuing to push a dead lead has a real cost: the guest's remaining window. The skill isn't persistence alone — it's knowing how much runway is left before a Plan B has to start, so that trying harder on Plan A doesn't consume the time needed to still deliver something on Plan B.
- Referral commissions are common and are not the problem — a recommendation distorted by one is. Trust is the entire product; a guest who later learns a "personal favorite" restaurant recommendation was a paid placement doesn't just distrust that recommendation, they distrust every recommendation that came before it.
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.
- 5d ago First seen · 89 lines · 83 tokens per session scan A 576ad6bb93fc
concierge is a skill published in the GitHub repository wonsukchoi/domain-experts (15 stars, last pushed yesterday), licensed MIT. It adds 83 tokens to every session and 2,461 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
planning-with-files
Persistent file-based planning for multi-step AI-agent work. Keeps taskplan.md, findings.md, and progress.md on disk; lifecycle hooks inject selected project planning context. Automatic recovery reads project planning files only. Explicit session-catchup.py --metadata reads same-project local agent session records and…
meeting-transcript
Process meeting recordings and notes into structured decisions, action items, and team dynamics with intelligent noise filtering.
outlook-calendar
A skill for managing Microsoft Outlook and Microsoft 365 calendar events through a connected account.
infrastructure-project
Skill for the project management infrastructure module providing multi-project discovery, structure validation, and metadata extraction. Use when discovering active projects, validating project directory structure, or extracting project configuration metadata.
infrastructure-rules
Skill for the rules module — discovery, validation, scope, and private-sidecar symlink sync for the top-level rules/ directory (specifications include soft markdown guidelines and strong yaml/json formal constraints). Use when discovering rules (discoverrules), resolving a rule path (resolveruleroot), validating rule…
infrastructure-tools
Skill for the tools module — discovery, validation, scope, and private-sidecar symlink sync for the top-level tools/ directory (executable entry points such as scripts, skills, and agents stored as passive manifests with scripts/ directories). Use when discovering tools (discovertools), resolving a tool path…