Agent skill

Restaurant Booking

by asgeirtj in asgeirtj/system_prompts_leaks

Find an available table and make, change, or cancel an authorized restaurant reservation.

CC0-1.0Auto-check passed

Install Restaurant Booking

skills CLI
$ npx skills add asgeirtj/system_prompts_leaks --skill restaurant-booking -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install asgeirtj/system_prompts_leaks restaurant-booking --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/asgeirtj/system_prompts_leaks.git skills-src && mkdir -p .claude/skills && cp -r skills-src/OpenAI/dots/skills/restaurant-booking .claude/skills/restaurant-booking && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
restaurant-booking
GitHub stars
69k
Token cost
~2.6k tokens
SKILL.md length
1,545 words
Files
1
Skills in repo
128
Repo updated
First seen
Licence
CC0-1.0

At a glance

Find an available table and make, change, or cancel an authorized restaurant reservation.

  • The user wants a booking handled
  • SKILL.md covers Before booking, Handle the reservation and Examples
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md
  • Not just restaurant ideas

What it does

Restaurant Booking is an agent skill from asgeirtj/system_prompts_leaks. Find an available table and make, change, or cancel an authorized restaurant reservation. Use when the user wants a booking handled, not just restaurant ideas.

Its SKILL.md is about 2.6k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

The repository describes itself as: Documented system prompts from Anthropic - Claude Fable 5.1, Opus 5.5, Claude Design, Claude Code. OpenAI - ChatGPT GPT-6-Astra, Codex. Google - Gemini 3.8 Flash, 3.1 Pro… The licence is CC0-1.0.

When your agent uses it

  • The user wants a booking handled
  • Not just restaurant ideas

Example prompts

  • “/restaurant-booking”

What it can do on your machine

Read from SKILL.md and the folder at commit 60d44cc. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Restaurant Booking loads about 2.6k tokens when it runs. Until then it costs about 45 tokens; SKILL.md has 1,545 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~45
When it runs · the whole SKILL.md, loaded when a task matches
~2.6k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from asgeirtj/system_prompts_leaks at commit 60d44cc, republished under its CC0-1.0 licence (© asgeirtj). 1,545 words, ~2,573 tokens.

Download SKILL.mdSave it as .claude/skills/restaurant-booking/SKILL.md (or your agent's skills folder).
name
restaurant-booking
description
Find an available table and make, change, or cancel an authorized restaurant reservation. Use when the user wants a booking handled, not just restaurant ideas.

Restaurant Booking

Get the right table and confirm what was actually booked.

Before booking

  • Start on OpenTable unless the user names another provider, or their known preferred or signed-in provider offers this restaurant. If the first provider can't book it, try the remaining options in order: OpenTable, Resy, then another booking provider verified through the restaurant's official site. Use the restaurant's direct flow only if no aggregator can book. Provider accounts are more likely to have the user's booking details and payment method already on file.

  • Check the venue and branch, local date and time range, party size, booking name, and seating or dietary needs. Look for an existing confirmation or later change before making a new reservation. Use $orbit:restaurant-recommendations if the restaurant still needs choosing.

  • Check what the booking has to fit: a show, commute, meeting, celebration, or another person's stated constraint. Use relevant past feedback about indoor or outdoor seating, noise, or accessibility without assuming it applies to everyone. If a missing fact changes the booking, first check existing notes or likely sources; ask only for what you still cannot establish.

  • If a booking provider such as Resy needs a number for texts, check relevant notes, past booking emails, or a profile before asking. Keep looking if the first source shows only part of the number; don't assume a work number is right for a personal booking. Finding a number does not authorize sharing it. Follow <confirmation_policy> for the specific number and provider before entering it. If it's likely to be the right one, ask, "Can I send your number ending in 0192 to hold the spot?" Ask for the full number only if you cannot find it.

  • Check live availability with the booking provider. If a read-only reservation-search tool is actually available, it can help find times; it cannot authenticate, pay, or book. Complete the reservation in the browser at the provider's exact restaurant, date, time, and party-size checkout, not the restaurant homepage. An open restaurant is not the same as an available table.

  • If the requested slot is full, check nearby times within the agreed range, the restaurant's published table-release time, and whether it offers a waitlist. Only join the waitlist when authorized. Say a monitor is set only after a supported tool confirms it.

  • If no requested slot is available, say what you checked and offer a verified alternative. Offer to watch for cancellations; set up a monitor only if the user agrees or already asked you to keep checking. Set a cutoff for when the reservation would no longer be useful.

    • Ex: "There's nothing for two Saturday between 6 and 8. Want me to watch for cancellations? Rintaro has 6:30 Friday if you're flexible."
    • If the user already said, "Keep checking and let me know if a table opens," set up the monitor without asking again and confirm when it will stop.
  • Read the terms for the specific table: deposits, prepaid menus, minimum spend, cancellation or no-show fees, and any time limit. Surface charges or restrictions the request doesn't cover before submitting.

Handle the reservation

  • Follow <confirmation_policy> when booking, changing, or cancelling. Keep the venue and time within the user's permission. If emailing the restaurant is the next step, use $orbit:email for the first outbound email and follow <confirmation_policy> before sending; a request to book does not by itself authorize a fallback email. Join a waitlist only when authorized, and report it as a waitlist until a table is confirmed.
  • Prefer an existing signed-in provider session and a card visibly saved there when the user's authorization covers its use. If sign-in is needed, request secure sign-in at the provider early, explain what the user needs to do, and resume promptly afterward. Don't send the user to the restaurant homepage or ask for fresh card details when the provider's saved card can complete the booking. Don't promise the session will stay signed in for future bookings.
  • Carry approval through the same booking or renewed temporary hold when the terms haven't materially changed; don't ask again just because sign-in or an expired hold interrupted it. Surface a newly introduced or materially changed payment, deposit, no-show fee, or cancellation term and follow <confirmation_policy>. If anything blocks you, tell the user rather than silently stopping.
  • For changes, check the replacement and the cancellation terms before giving up the original table. If submission gives an unclear result, check the reservation or confirmation before retrying.
  • Share dietary or accessibility requirements accurately and only as authorized. An identifiable person's allergy or health information requires authorization for that specific information and destination under <confirmation_policy>; ask a general question without naming the person when that is enough. A booking note does not prove the kitchen can avoid cross-contact. If no usable provider-stored card is available, use the Wallet skill when supported and authorized; calendar changes and messages to companions follow <confirmation_policy> too.
  • Return the restaurant and address, local time, party size, confirmation, and any fee, cancellation deadline, or request still unconfirmed.
  • For an important meal, include arrival instructions or the deadline that would cost the user money, if relevant. If an authorized cancellation or change has a fee, verify the provider's actual outcome and expected refund instead of assuming it went through.
Show full SKILL.md (724 more words)Show less

Examples

1. The requested table is full

  • User: "Can you book a table for four at Laundromat at 8 on Sunday?"
  • Action: Confirm both booking routes have no availability, check for a waitlist, and verify The Pearl has a table for four at 7:30.
  • Guidance: Offer a verified alternative. Ask before joining a waitlist unless the user already authorized it; a waitlist is not a booking.
  • dot: "Laundromat is full for four at 8 Sunday. The Pearl has a 7:30 spot for 4 and is just a few streets away. Should I join Laundromat's waitlist for you, or snag the Pearl spot?"

2. A birthday dinner after a movie

  • User: "I want 6 of us to do a birthday dinner for Sarah at Penny Roma after our 5pm Odyssey screening."
  • Action: Use the 5 p.m. screening to look for a table around 8.
  • Action: Verify the 8:15 table for six. Proactively look into the cake policy, and the card-hold terms before asking to book.
  • Guidance: Leave time after the movie. Verify outside-cake rules and disclose any cancellation fee or deposit before asking the user to decide.
  • dot: "Got it. Penny Roma has a table for six at 8:15 which should give you enough time after your movie. You can even bring Sarah's cake for free. They just need a card to hold it - want me to do that?"

3. A team dinner with step-free access

  • User: "Find us a table for a 15-person team dinner. We need step-free access."
  • Action: Check the restaurant's official accessibility information and where the table is. Verify how large parties book; use $orbit:email only if authorized.
  • Guidance: A general accessibility question needn't identify anyone. A listed ramp and table don't prove a reservation is available.
  • dot: "Flour and Water lists a ramp and a table for 15. They book groups by email. Want me to ask?"

4. A later table has different terms

  • User: "Move our 7 p.m. dinner to 8."
  • Action: Verify that 8 is full and 8:15 is available with a prepaid menu. Keep the 7 p.m. booking while the user decides.
  • Guidance: Don't give up the original reservation until the replacement and cancellation terms are clear. Ask before accepting an unapproved prepaid menu.
  • dot: "8 is full. 8:15 is open but requires a prepaid menu. I'm keeping your 7 p.m. booking for now. Do you want me to switch it?"

5. Resy needs a mobile number

  • User: "Book dinner for two at 7 on Saturday at The Progress, on Resy."

  • Action: For The Progress for two at 7 on Saturday, first check prior bookings and relevant records for the number; verify the table is available and Resy requires it. If authorization is missing, ask before entering it.

  • Guidance: A booking request doesn't by itself authorize sharing a private number. If this specific disclosure to Resy was already authorized, don't ask again. Show only enough of the number for the user to recognize it, and ask only for what remains uncertain.

  • dot: "The Progress has a table for two at 7 on Saturday. Can I send your number ending in 0192 to hold the spot?"

  • User: "Yup."

  • Action: Finish and verify the reservation, then send the confirmation. Offer to add it to the calendar unless applicable standing permission already covers that step; if it does, add it, verify it, and report it.

  • dot:

    text
    "The Progress is booked for two at 7 on Saturday. Here's your [confirmation](LINK_URL). Should I add it to your calendar?"

6. Sign in once and resume the same booking

  • User: "Book The Progress for four around 7."

  • Action: Establish the intended local date from context, or ask if it is missing, before opening the exact provider checkout. If sign-in is needed, ask early and link the secure OpenTable sign-in; don't ask for fresh card details.

  • Guidance: Carry approval through the same unchanged booking or temporary hold. Follow <confirmation_policy> if a fee or other material term changes, and don't call it booked until OpenTable confirms it.

  • dot:

    text
    "Please sign in to [OpenTable](LINK_URL), then I'll finish booking with your saved card!"
  • User: "Done."

  • Action: Resume promptly without asking again if the terms haven't materially changed. Verify the reservation, then share its confirmed local time and confirmation.

  • dot (only after provider confirmation):

    text
    "Booked The Progress for four at [confirmed time]. Here's your [confirmation](LINK_URL). Enjoy!"

© asgeirtj, CC0-1.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in OpenAI/dots/skills/restaurant-booking of asgeirtj/system_prompts_leaks.

Open the folder on GitHubat commit 60d44cc

Compare with similar skills

Restaurant Booking next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Restaurant Booking compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Restaurant Booking this skillasgeirtj/system_prompts_leaks69k—~2.6kAutomated safety check: PassCC0-1.0
Restaurant Bookinggooseworks-ai/goose-skills1.2k1 repos~1.7kAutomated safety check: PassMIT
Hermes Agent Skill AuthoringNousResearch/hermes-agent252k—~3.6kAutomated safety check: PassMIT
Configuring Oauth2 Authorization Flowmukul975/Anthropic-Cybersecurity-Skills34k—~1.7kAutomated safety check: PassApache-2.0
BookingsBuilderIO/agent-native7.1k—~345Automated safety check: PassNone
Authoring Skillsvercel/next.js143k—~1kAutomated safety check: PassMIT

Similar skills

  • Restaurant Booking

    gooseworks-ai/goose-skills

    Book restaurant reservations via browser automation. An agent skill from gooseworks-ai/goose-skills.

    1.2k GitHub starsUsed in 1 repo~1.7k tokens
    Productivity & AutomationAuto-check passed
  • Hermes Agent Skill Authoring

    NousResearch/hermes-agent

    Author in-repo SKILL.md files: frontmatter and structure. An agent skill from NousResearch/hermes-agent.

    252k GitHub stars~3.6k tokensUpdated yesterday
    Agent WorkflowsAuto-check passed
  • Configuring Oauth2 Authorization Flow

    mukul975/Anthropic-Cybersecurity-Skills

    Configures secure OAuth 2.0 authorization flows, including Authorization Code with PKCE, Client Credentials, and Device Authorization Grant, covering flow selection, PKCE implementation, token…

    34k GitHub stars~1.7k tokensUpdated 1 mo ago
    Backend & APIsAuto-check passed
  • Bookings

    BuilderIO/agent-native

    Booking lifecycle — pending, confirmed, rescheduled, cancelled — plus attendees, references, no-shows, and reminders.

    7.1k GitHub stars~345 tokensUpdated yesterday
    Auto-check passed
  • Authoring Skills

    vercel/next.js

    Official

    How to create and maintain agent skills in .agents/skills/. An agent skill from vercel/next.js.

    143k GitHub stars~1k tokensUpdated today
    Agent WorkflowsAuto-check passed
  • Book to Skill Converter

    virgiliojr94/book-to-skill

    Converts books and documents in PDF, EPUB, DOCX, HTML, Markdown, text, RTF or MOBI form into agent skills built from frameworks, principles, techniques and anti-patterns.

    34k GitHub stars~14k tokensUpdated 5 days ago
    Knowledge ManagementAuto-check passed

More from asgeirtj/system_prompts_leaks

All 125 skills in this repo
  • Fleet Manager for Agent Sessions

    asgeirtj/system_prompts_leaks

    Shows one digest of coding-agent sessions across your connected machines and lets you open, read, steer, approve, stop and close them, over Herdr, tmux or MSP.

    69k GitHub stars~2.5k tokensUpdated today
    Auto-check passed
  • Muse Code Product Doctor

    asgeirtj/system_prompts_leaks

    Diagnoses a Muse Code installation's own failures from binary and session evidence, instead of treating the report as an ordinary repository bug.

    69k GitHub stars~3.5k tokensUpdated today
    Auto-check passed
  • DOCX

    asgeirtj/system_prompts_leaks

    A skill your agent uses whenever the user wants to create, read, edit, or manipulate Word documents (.docx) or Word templates (.dotx).

    69k GitHub stars~1.9k tokensUpdated today
    Auto-check passed
  • Agents Project Coordinator

    asgeirtj/system_prompts_leaks

    Runs a goal as a project in which the agent coordinates separate agent threads, judging when to split the work, and interviews you first when nothing can be verified.

    69k GitHub stars~2.9k tokensUpdated today
    Auto-check passed
  • Muse Plugin Creator

    asgeirtj/system_prompts_leaks

    Creates and validates a new native Muse plugin package in the current workspace, limited to five capability families, and leaves installation to you.

    69k GitHub stars~1.8k tokensUpdated today
    Auto-check passed
  • Deep Research

    asgeirtj/system_prompts_leaks

    A skill your agent uses when the user's prompt requires (1) researching a topic across multiple sources, comparing options or alternatives, analyzing trends or history, understanding markets or…

    69k GitHub stars~3.3k tokensUpdated today
    Auto-check passed

Questions about Restaurant Booking

What does Restaurant Booking do?

Find an available table and make, change, or cancel an authorized restaurant reservation. Restaurant Booking is an agent skill from asgeirtj/system_prompts_leaks. Find an available table and make, change, or cancel an authorized restaurant reservation.

When should I use Restaurant Booking?

Restaurant Booking fits situations like: the user wants a booking handled; not just restaurant ideas.

How do I install Restaurant Booking in Claude Code?

Run `npx skills add asgeirtj/system_prompts_leaks --skill restaurant-booking -a claude-code`. Or copy the skill folder (OpenAI/dots/skills/restaurant-booking in asgeirtj/system_prompts_leaks) into .claude/skills/restaurant-booking in your project. Claude Code loads it when a task matches its description.

How do I install Restaurant Booking in Codex?

Run `npx skills add asgeirtj/system_prompts_leaks --skill restaurant-booking -a codex`. Or copy the skill folder (OpenAI/dots/skills/restaurant-booking in asgeirtj/system_prompts_leaks) into .agents/skills/restaurant-booking in your project. Codex loads it when a task matches its description.

Can I use Restaurant Booking in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add asgeirtj/system_prompts_leaks --skill restaurant-booking -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/restaurant-booking, .gemini/skills/restaurant-booking, .github/skills/restaurant-booking and .opencode/skills/restaurant-booking in your project.

What does Restaurant Booking need to run?

SKILL.md names no scripts, command-line tools or credentials: Restaurant Booking is instructions for the agent only.

Does Restaurant Booking access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Restaurant Booking safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Restaurant Booking use?

Restaurant Booking is published under the CC0-1.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Restaurant Booking use?

About 2.6k tokens (SKILL.md is roughly 10k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Restaurant Booking?

Skills that share tags, products or a category with Restaurant Booking: Restaurant Booking (gooseworks-ai/goose-skills, 1.2k stars), Hermes Agent Skill Authoring (NousResearch/hermes-agent, 252k stars), Configuring Oauth2 Authorization Flow (mukul975/Anthropic-Cybersecurity-Skills, 34k stars) and Bookings (BuilderIO/agent-native, 7.1k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Restaurant Booking?

asgeirtj (a GitHub user) maintains it in asgeirtj/system_prompts_leaks, which has 69,280 GitHub stars. The repository holds 128 skills in this directory. The repository was last updated on October 10, 2026.

Source: asgeirtj/system_prompts_leaks on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.