← 返回 Skills 市场
454
总下载
0
收藏
0
当前安装
2
版本数
在 OpenClaw 中安装
/install swarm-safety
功能描述
SWARM: System-Wide Assessment of Risk in Multi-agent systems. 38 agent types, 29 governance levers, 55 scenarios. Study emergent risks, phase transitions, an...
安全使用建议
This skill appears coherent with its stated purpose, but standard caution is advised: (1) confirm the PyPI package name and the GitHub repo (https://github.com/swarm-ai-safety/swarm) match what you expect before installing; (2) install in a virtual environment or sandbox; (3) do not bind the API to 0.0.0.0 or expose it to untrusted networks; (4) do not include real API keys, passwords, or PII in scenario YAMLs; (5) if you need higher assurance, review the package source on GitHub or the PyPI release contents and checksums before installing.
功能分析
Type: OpenClaw Skill
Name: swarm-safety
Version: 1.7.1
The OpenClaw AgentSkills skill bundle for 'swarm-safety' appears benign. It describes a research framework for AI safety, explicitly outlining security measures such as API binding to localhost only and restricting CORS. The documentation includes 'Hard Rules' and 'Security Notes' that guide the AI agent towards secure behavior, such as not submitting real API keys or PII, and warning against exposing the development API to untrusted networks. Installation and usage instructions are standard for a Python package and CLI tool, with no evidence of data exfiltration, malicious execution, persistence mechanisms, or prompt injection attempts designed to subvert the agent's security or privacy.
能力评估
Purpose & Capability
Name/description (multi-agent safety research) align with the instructions (pip-installable Python package, simulation API, CLI). The skill does not request unrelated resources (no cloud creds, no unusual binaries).
Instruction Scope
SKILL.md focuses on running local simulations, CLI usage, and a localhost-only API. It explicitly warns not to expose the dev API, not to include real credentials/PII in scenarios, and treats results as research artifacts. The instructions do not direct the agent to read unrelated system files or exfiltrate data.
Install Mechanism
No install spec in registry (instruction-only). SKILL.md recommends standard pip installs or cloning the GitHub repo — common, well-understood install methods with no opaque download URLs.
Credentials
The skill declares no required environment variables, credentials, or config paths. The SKILL.md does show that agent registration returns an api_key for local API use — appropriate and proportional to the stated API functionality.
Persistence & Privilege
always is false and the skill is user-invocable. The SKILL.md says storage is in-memory by default and warns about production deployment changes; it does not request permanent agent-level privileges or modify other skills.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install swarm-safety - 安装完成后，直接呼叫该 Skill 的名称或使用
/swarm-safety触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.7.1
Trigger sync to openclaw/skills GitHub repo
v1.7.0
**Expanded research capabilities and scenario coverage in multi-agent AI safety simulation.**
- Increased to 38 agent types, 29 governance levers, and 55 scenarios in v1.7.0.
- Enhanced documentation with installation, usage (Python, CLI, API), and security guidelines.
- Detailed explanation of key metrics and core mechanisms for studying emergent risks.
- Added clear when-to-use guidance and hard operational rules for research context.
- Outlined all supported agent families and governance mechanisms.
元数据
常见问题
SWARM Safety 是什么？
SWARM: System-Wide Assessment of Risk in Multi-agent systems. 38 agent types, 29 governance levers, 55 scenarios. Study emergent risks, phase transitions, an... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 454 次。
如何安装 SWARM Safety？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install swarm-safety」即可一键安装，无需额外配置。
SWARM Safety 是免费的吗？
是的，SWARM Safety 完全免费（开源免费），可自由下载、安装和使用。
SWARM Safety 支持哪些平台？
SWARM Safety 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 SWARM Safety？
由 rsavitt（@rsavitt）开发并维护，当前版本 v1.7.1。
推荐 Skills