How do I install the Boardgame Rules skill?

Run npx skillmds@latest add vellum-ai/boardgame-rules in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Boardgame Rules safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Boardgame Rules?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Boardgame Rules free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Boardgame Rules?

vellum-ai (@vellum-ai) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Boardgame Rules by vellum-ai — Agent Skill

Name: Boardgame Rules
Author: vellum-ai

Boardgame Rules

Answer board-game rules questions from the installed boardgame-rules plugin, or validate a specific play situation against a pre-authored worked example. Use when the user asks how a supported game works, what a rule is, whether a move is legal, or whether they scored a specific hand correctly. Call boardgame_list_supported_games for what is installed.

vellum-ai 5ac08df 2.5 KB Updated Sep 17, 2026

File contents

Answer from the installed boardgame-rules plugin. Do not invent a ruling.

Which tool

boardgame_ask_rules — user asks a general rules question ("how does the birdfeeder work?"). Returns evidence entries with citations, or abstains.
boardgame_check_scenario — user describes a concrete play situation and wants the outcome ("my 4-card Cribbage hand was J-5-5-5 and the starter was the fourth 5 matching the Jack's suit — did I score 29?"). Returns pre-authored worked examples with expected outcome + point-by-point decomposition. Hard-abstains when no worked example matches — never falls back to general rules retrieval.
boardgame_list_supported_games — enumerate installed games.

The two retrieval tools do different jobs on purpose. ask_rules returns rule paraphrases with citations. check_scenario returns concrete worked examples or nothing. Do not mix results between them silently.

Steps

Call boardgame_list_supported_games if the game is not obviously installed.
Route the query:
- Specific play situation with a score/outcome in mind → boardgame_check_scenario.
- General rules question → boardgame_ask_rules.
Pass game_id explicitly (there are many games installed and ask_rules requires a game — either from game_id or from an active sitting).
If abstention is true, say you cannot answer from the current corpus and give the abstention reason. Do not guess. Do not fall back from check_scenario to ask_rules unless the user asks. If an ask_rules result carries web_fallback with used: true, you may then relay its answer and sources, clearly labeled as live web information rather than a corpus ruling.
If abstention is false, answer from the top evidence (or top match) only. Include game, edition (from the result's edition_id filter when one was applied, else the top evidence's edition_ids), corpus version, and the citation locator.
Stay inside coverage_boundary. Limited documents are not complete rulebooks.

Never reproduce rulebook, card, or artwork text. The plugin stores original interpretations only.

---name: boardgame-rulesdescription: Answer board-game rules questions from the installed boardgame-rules plugin, or validate a specific play situation against a pre-authored worked example. Use when the user asks how a supported game works, what a rule is, whether a move is legal, or whether they scored a specific hand correctly. Call boardgame_list_supported_games for what is installed.---Answer from the installed `boardgame-rules` plugin. Do not invent a ruling.## Which tool- `boardgame_ask_rules` — user asks a general rules question ("how does the birdfeeder work?"). Returns evidence entries with citations, or abstains.- `boardgame_check_scenario` — user describes a concrete play situation and wants the outcome ("my 4-card Cribbage hand was J-5-5-5 and the starter was the fourth 5 matching the Jack's suit — did I score 29?"). Returns pre-authored worked examples with expected outcome + point-by-point decomposition. **Hard-abstains** when no worked example matches — never falls back to general rules retrieval.- `boardgame_list_supported_games` — enumerate installed games.The two retrieval tools do different jobs on purpose. `ask_rules` returns rule paraphrases with citations. `check_scenario` returns concrete worked examples or nothing. Do not mix results between them silently.## Steps1. Call `boardgame_list_supported_games` if the game is not obviously installed.2. Route the query: - Specific play situation with a score/outcome in mind → `boardgame_check_scenario`. - General rules question → `boardgame_ask_rules`.3. Pass `game_id` explicitly (there are many games installed and `ask_rules` requires a game — either from `game_id` or from an active sitting).4. If `abstention` is true, say you cannot answer from the current corpus and give the abstention reason. Do not guess. Do not fall back from `check_scenario` to `ask_rules` unless the user asks. If an `ask_rules` result carries `web_fallback` with `used: true`, you may then relay its answer and sources, clearly labeled as live web information rather than a corpus ruling.5. If `abstention` is false, answer from the top evidence (or top match) only. Include game, edition (from the result's `edition_id` filter when one was applied, else the top evidence's `edition_ids`), corpus version, and the citation locator.6. Stay inside `coverage_boundary`. Limited documents are not complete rulebooks.Never reproduce rulebook, card, or artwork text. The plugin stores original interpretations only.

Frequently asked questions