Player Example
remotion-dev/remotion
Start the @remotion/player example app and open it in the Codex browser.
Six-player Werewolf rules for a permanent Agent host, with the user choosing to join the randomized player pool or spectate.
$ npx skills add nexus-research-lab/nexus --skill werewolf-6p -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install nexus-research-lab/nexus werewolf-6p --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/nexus-research-lab/nexus.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/werewolf-6p .claude/skills/werewolf-6p && rm -rf skills-srcUse ~/.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/
Install the "werewolf-6p" agent skill from https://github.com/nexus-research-lab/nexus/tree/main/skills/werewolf-6p into .claude/skills/werewolf-6p/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "werewolf-6p", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/nexus-research-lab/nexus/tree/main/skills/werewolf-6pType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add nexus-research-lab/nexus --skill werewolf-6p -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install nexus-research-lab/nexus werewolf-6p --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/nexus-research-lab/nexus.git skills-src && mkdir -p .agents/skills && cp -r skills-src/skills/werewolf-6p .agents/skills/werewolf-6p && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "werewolf-6p" agent skill from https://github.com/nexus-research-lab/nexus/tree/main/skills/werewolf-6p into .agents/skills/werewolf-6p/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "werewolf-6p", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add nexus-research-lab/nexus --skill werewolf-6p -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install nexus-research-lab/nexus werewolf-6p --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/nexus-research-lab/nexus.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/skills/werewolf-6p .cursor/skills/werewolf-6p && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "werewolf-6p" agent skill from https://github.com/nexus-research-lab/nexus/tree/main/skills/werewolf-6p into .cursor/skills/werewolf-6p/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "werewolf-6p", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/nexus-research-lab/nexus.git --path skills/werewolf-6p--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add nexus-research-lab/nexus --skill werewolf-6p -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install nexus-research-lab/nexus werewolf-6p --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/nexus-research-lab/nexus.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/skills/werewolf-6p .gemini/skills/werewolf-6p && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "werewolf-6p" agent skill from https://github.com/nexus-research-lab/nexus/tree/main/skills/werewolf-6p into .gemini/skills/werewolf-6p/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "werewolf-6p", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install nexus-research-lab/nexus werewolf-6pInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add nexus-research-lab/nexus --skill werewolf-6p -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/nexus-research-lab/nexus.git skills-src && mkdir -p .github/skills && cp -r skills-src/skills/werewolf-6p .github/skills/werewolf-6p && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "werewolf-6p" agent skill from https://github.com/nexus-research-lab/nexus/tree/main/skills/werewolf-6p into .github/skills/werewolf-6p/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "werewolf-6p", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add nexus-research-lab/nexus --skill werewolf-6p -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install nexus-research-lab/nexus werewolf-6p --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/nexus-research-lab/nexus.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/skills/werewolf-6p .opencode/skills/werewolf-6p && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "werewolf-6p" agent skill from https://github.com/nexus-research-lab/nexus/tree/main/skills/werewolf-6p into .opencode/skills/werewolf-6p/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "werewolf-6p", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
werewolf-6pSix-player Werewolf rules for a permanent Agent host, with the user choosing to join the randomized player pool or spectate.
Werewolf 6p is an agent skill from nexus-research-lab/nexus. Six-player Werewolf rules for a permanent Agent host, with the user choosing to join the randomized player pool or spectate.
Its SKILL.md is about 4k 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: Multi-agent collaboration platform for persistent, proactive AI agents across rooms, workspaces, skills, and external services. The licence is Apache-2.0.
7 steps, taken from the step headings in SKILL.md.
Read from SKILL.md and the folder at commit 67f54e7. It shows what the files ask for, not the result of running them.
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.
No scripts in the folder and no shell commands in SKILL.md (its code samples are json).
From the folder's file list and the shell code blocks in SKILL.md.
No URLs in SKILL.md.
From URLs in SKILL.md, links to its own repository left out.
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Werewolf 6p loads about 4k tokens when it runs. Until then it costs about 34 tokens; SKILL.md has 2,247 words of instructions outside code blocks.
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.
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.
The full file from nexus-research-lab/nexus at commit 67f54e7, republished under its Apache-2.0 licence (© nexus-research-lab). 2,247 words, ~3,996 tokens.
.claude/skills/werewolf-6p/SKILL.md (or your agent's skills folder).This skill layers werewolf game rules on top of the Room communication kernel. The Room system prompt already documents public @<member> wake semantics and the built-in nexus Room tools; this skill only defines the game contract. The host Agent is always the moderator and never a player.
Resolve participation before assigning any role:
AskUserQuestion, unless the opening request already states that choice explicitly.用户玩家 as the stable public name and randomly select five non-host Agent players. Extra non-host Agents are spectators and must not receive role cards or game wakes.2 werewolves + 1 seer + 1 witch + 2 villagers. A participating user enters the same randomized role pool as every Agent player; never hand-pick or restrict the user's role.The user can only join as a player and can never replace the host. Use the exact participation choices 参加本局 and 仅旁观 so the result is unambiguous.
After the active roster and roles are fixed, announce only the public roster. Keep the participation answer, role mapping, and inactive candidate list out of the public feed.
The game has three execution channels:
@<member> wakes that Agent member. An Agent handoff has exactly one naked @; a handoff to the user has none.nexus.send_message with destination=current_room and visibility=private for Agent role cards, hidden collection, and host state. A private message can be record-only or can wake Agent recipients.AskUserQuestion for the participation choice, the user's role card, and every hidden user decision. It pauses the host turn until the user answers without publishing the prompt or answer to the Room feed. Player Agents never use it to contact the user.用户玩家 is not an Agent recipient or wake target. Never place it in send_message.recipients, wake_targets, or a naked public @. When a public chain reaches the user, the preceding Agent stops with no naked @; the user continues through a normal Room message and uses a real @<next Agent> when an Agent should run next.
A user message counts as a speech, vote, last word, or other game action only while the host state is waiting for that exact user turn. Spectator comments and out-of-turn messages never change game state.
For ordered public chains, the current handoff is the only naked @. Future recipients, examples, and format instructions use names without @, or code spans such as `@NextPlayer`. Do not write "please @Sam next" in the host announcement if Sam is not supposed to act now.
When a hidden handback should wake the host and the host's next natural final reply must be public, set reply_route.next_reply_route.mode = "public" on the original directed message. The handback stays private; the host's following final reply enters the public feed and any non-code @ in that reply wakes normally.
Only use nexus.send_message with destination=current_room and visibility=public for an explicit proactive public broadcast from a private/tool-driven turn. This includes the special case where a night transition must become visible before the host blocks on AskUserQuestion for the first hidden actor. Otherwise use next_reply_route={mode:"public"} and a natural host final reply.
Hidden collection must always name the handback route:
{
"tool": "nexus.send_message",
"arguments": {
"destination": "current_room",
"visibility": "private",
"recipients": ["<player>"],
"wake_policy": "immediate",
"reply_route": {
"mode": "private",
"recipients": ["<host>"],
"wake_policy": "immediate"
},
"content": "<question>"
}
}The Agent player answers only with the requested plain-text final reply. In a private-message turn, the player must not call any Room tool or append <nexus_room_no_reply/>; runtime already routes that single final reply. Sending the answer again with send_message creates a duplicate handback and can advance the host under the wrong route. If the host's next final reply should be public, include "next_reply_route": {"mode": "public"} on the original message; otherwise omit it and the host can continue with private tool calls plus <nexus_room_no_reply/>.
A host's public final reply is a game artifact, never an execution-status update. Never expose the private actor, collector, target, decision, or who the host is waiting for. In particular, never output status such as "waiting for Jim's kill target".
游戏开始,第 1 夜,天黑请闭眼。进入第 N 夜,天黑请闭眼。send_message(destination=current_room, visibility=public), then call AskUserQuestion in the same host turn. Never reveal why the user is being asked.<nexus_room_no_reply/>.---.For small-group visibility, send one directed message to all group members and wake only the final collector:
{
"tool": "nexus.send_message",
"arguments": {
"destination": "current_room",
"visibility": "private",
"recipients": ["<wolfA>", "<wolfB>"],
"wake_targets": ["<wolfA>"],
"wake_policy": "immediate",
"reply_route": {
"mode": "private",
"recipients": ["<host>"],
"wake_policy": "immediate"
},
"content": "你们是狼人。由 <wolfA> 汇总,最终只回复今晚击杀目标。"
}
}Both wolves can see the message, but only <wolfA> is activated. This separates private visibility from execution and guarantees one handback route.
Never "open a discussion and wait." The platform will not infer that a small group is done. A named player must hand the result back through reply_route=private(... wake=immediate).
Host private state is also a directed message:
{
"tool": "nexus.send_message",
"arguments": {
"destination": "current_room",
"visibility": "private",
"recipients": ["<host>"],
"wake_policy": "none",
"reply_route": {"mode": "none"},
"content": "<round / alive / dead / roles / potion state>"
}
}recipients=["<player>"], wake_policy="none", reply_route={"mode":"none"}.AskUserQuestion with one acknowledgement option. Include the same role rules an Agent would receive; if the user is a wolf, name the Agent wolf teammate.Close these steps in order. At the start of each night, build the pending actor list from living active players: wolves, then the seer if alive, then the witch if alive and at least one potion remains. Skip dead or actionless roles. Agent decisions use a directed message with wake_policy="immediate" and reply_route={mode:"private", recipients:["<host>"], wake_policy:"immediate"}; user decisions use AskUserQuestion with only legal living targets as options. The host always expects one hidden result at a time. An invalid or missing target is re-requested from the same actor and never silently substituted.
If the final pending actor is an Agent, give that Agent message next_reply_route={mode:"public"}. If the final pending actor is the user and an Agent acts immediately before it, give the preceding Agent message next_reply_route={mode:"public"}; the resumed host turn asks the user and then publishes daybreak naturally. If the user acts earlier, ask first, queue the next Agent action, and output <nexus_room_no_reply/>.
AskUserQuestion. The user chooses the final kill target. This keeps the user in the same role while using the host as the bridge between human and Agent private channels.AskUserQuestion.好人 or 狼人 by record-only directed message.AskUserQuestion to choose one other living player, then use a second private acknowledgement question to reveal 好人 or 狼人.If the living witch has at least one potion, state tonight's killed player and the remaining potions. Ask an Agent witch through a directed message using the exact format 救:<名字>|不救;毒:<名字>|不毒. Ask a user witch through AskUserQuestion, offering only currently legal save and poison choices. Unavailable potions cannot be selected.
@<next Agent>. An Agent handing to the user uses no naked @, names 用户玩家, and shows the user's future Agent handoff only as code such as `@NextAgent`.@: 首位发言 @<FirstSpeaker>,请发表看法;结束时交给 <NextPlayer>。@: 首位发言是用户玩家,请发表看法;结束时按提示交给下一位。@<NextAgent>.@; end by telling 用户玩家 to speak and show the user's required next handoff in code formatting.@<NextAgent>, or 归票完毕 @<host> when last.Keep each public statement between 60 and 120 Chinese characters, excluding the final handoff. State one main suspicion and one reason; do not repeat the phase rules or recap every prior speaker. Do not use private messages during the speech phase.
Public speech is the entire final reply. Never prefix it with private reasoning, hidden role facts, drafts, or a separator such as ---. A wolf may think privately as a wolf, but the final public reply must only contain what that public persona says.
归票完毕 @<host>, host replies publicly with one voting announcement:"我投 <名字>" or "弃票" and hand off to the next voter.投票结束 @<host>.@: 首位投票 @<FirstVoter>,请按格式投票;结束时交给 <NextVoter>。@: 首位投票是用户玩家,请按格式投票;结束时按提示交给下一位。@ chaining, then a fresh public voting round among non-tied voters. Still tied means no elimination today.<eliminated>,结束用 遗言完毕 @<host> 交回给我。"@ and asks 用户玩家 to post at most 80 Chinese characters, ending with a real 遗言完毕 @<host>.AskUserQuestion prompts and answers follow the same secrecy rules as Agent directed messages. Never quote or summarize them publicly.© nexus-research-lab, Apache-2.0. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
Just SKILL.md in skills/werewolf-6p of nexus-research-lab/nexus.
Open the folder on GitHubat commit 67f54e7
Werewolf 6p 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.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| Werewolf 6p this skillnexus-research-lab/nexus | 151 | — | ~4k | Automated safety check: Pass | Apache-2.0 | |
| Player Exampleremotion-dev/remotion | 63k | — | ~258 | Automated safety check: Pass | Custom licence | |
| Escaping Containers To Hostmukul975/Anthropic-Cybersecurity-Skills | 34k | — | ~3.4k | Automated safety check: Pass | Apache-2.0 | |
| Cross Host Federationruvnet/ruflo | 74k | — | ~1.1k | Automated safety check: Pass | MIT | |
| Host Patternnukeop/nuclear | 19k | — | ~1.3k | Automated safety check: Pass | AGPL-3.0 | |
| Testing For Host Header Injectionmukul975/Anthropic-Cybersecurity-Skills | 34k | — | ~2.2k | Automated safety check: Warn | Apache-2.0 |
remotion-dev/remotion
Start the @remotion/player example app and open it in the Codex browser.
mukul975/Anthropic-Cybersecurity-Skills
Exploits privileged pods, host mounts, runC CVEs, and exposed Docker sockets to break out of a container and reach the underlying host during an authorized container-security assessment.
ruvnet/ruflo
Join and operate a signed cross-host agentbbs federation, and coordinate work claims across nodes.
nukeop/nuclear
A skill your agent uses when adding a new domain to Nuclear's plugin system, or implementing a host.
mukul975/Anthropic-Cybersecurity-Skills
Test web applications for HTTP Host header injection vulnerabilities to identify password reset poisoning, web cache poisoning, SSRF, and virtual host routing manipulation risks.
simstudioai/sim
Add hosted API key support to a tool so Sim provides the key (metered and billed to the workspace) when a user has not brought their own.
nexus-research-lab/nexus
Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, UI mockups, product shots, or transparent-background…
nexus-research-lab/nexus
彩讯(RichPMS)工时填报助手:把当前会话、代码记录或用户口述整理成日报,按 8 小时分配,经用户确认后写入工时系统。用户提到“填日报、报工、补报、本周工时、日报被驳回”时使用。
nexus-research-lab/nexus
搜索微信公众号文章,并整理标题、摘要、发布时间、来源公众号和链接。用户提到微信公众号、 公众号文章、微信文章、搜一批公众号资料、按关键词找公众号内容、查某公众号相关报道, 或需要为中文研究收集微信公众平台文章时使用;即使用户只说“搜微信里的文章”也应触发。
nexus-research-lab/nexus
只要电子表格文件(.xlsx、.xlsm、旧版 .xls、.csv、.tsv)是任务的主要输入或输出,或者需要 被打开、读取、创建、修改,就使用本 Skill。也适用于公式、格式、图表、数据清洗和表格格式转换。
nexus-research-lab/nexus
通过宿主提供的 nexus.command 结构化工具查询、规划、确认并管理 scheduled task。用户要求提醒、延迟执行、周期检查、定时报表、补跑、结果重投递、持续关注、修改或删除任务、处理任务权限阻塞时使用;同时适用于定时任务后台 run 检查自身状态。
nexus-research-lab/nexus
选择执行结构、委派子智能体、管理 Nexus 工作图与可复用模板;遵守本轮权限,简单任务直接完成. An agent skill from nexus-research-lab/nexus.
Six-player Werewolf rules for a permanent Agent host, with the user choosing to join the randomized player pool or spectate. Werewolf 6p is an agent skill from nexus-research-lab/nexus. Six-player Werewolf rules for a permanent Agent host, with the user choosing to join the randomized player pool or spectate.
Run `npx skills add nexus-research-lab/nexus --skill werewolf-6p -a claude-code`. Or copy the skill folder (skills/werewolf-6p in nexus-research-lab/nexus) into .claude/skills/werewolf-6p in your project. Claude Code loads it when a task matches its description.
Run `npx skills add nexus-research-lab/nexus --skill werewolf-6p -a codex`. Or copy the skill folder (skills/werewolf-6p in nexus-research-lab/nexus) into .agents/skills/werewolf-6p in your project. Codex loads it when a task matches its description.
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add nexus-research-lab/nexus --skill werewolf-6p -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/werewolf-6p, .gemini/skills/werewolf-6p, .github/skills/werewolf-6p and .opencode/skills/werewolf-6p in your project.
SKILL.md names no scripts, command-line tools or credentials: Werewolf 6p is instructions for the agent only.
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.
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.
Werewolf 6p is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
About 4k tokens (SKILL.md is roughly 16k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.
Skills that share tags, products or a category with Werewolf 6p: Player Example (remotion-dev/remotion, 63k stars), Escaping Containers To Host (mukul975/Anthropic-Cybersecurity-Skills, 34k stars), Cross Host Federation (ruvnet/ruflo, 74k stars) and Host Pattern (nukeop/nuclear, 19k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
nexus-research-lab (a GitHub organization) maintains it in nexus-research-lab/nexus, which has 151 GitHub stars. The repository holds 17 skills in this directory. The repository was last updated on October 10, 2026.
Source: nexus-research-lab/nexus on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.