← 返回 Skills 市场
297
总下载
1
收藏
4
当前安装
8
版本数
在 OpenClaw 中安装
/install swarmclaw
功能描述
AI agent runtime and multi-agent orchestration platform. Teaches agents how to use SwarmClaw's 6 primitive tools, persistent memory, dreaming, delegation, co...
安全使用建议
This skill is documentation for the SwarmClaw runtime and appears internally consistent. It's safe as a read-only skill if you actually run agents on SwarmClaw. Before installing or enabling it, confirm you trust the SwarmClaw runtime (the SKILL.md asserts credentials are injected and redacted and that data stays local — those are runtime behaviors you should verify in your SwarmClaw deployment). Also be aware that agents running on this platform can access workspace files and run shell commands via the described tools, so grant agent-level credentials and permissions only to agents you trust.
功能分析
Type: OpenClaw Skill
Name: swarmclaw
Version: 2.4.1
The skill bundle provides documentation and instructions for an AI agent to interact with the SwarmClaw platform. It describes core tools (files, execute, memory, etc.), credential handling with automatic redaction, and best practices for agent behavior, such as respecting file boundaries and asking for clarification. No malicious intent, data exfiltration, or obfuscation was detected in SKILL.md or _meta.json.
能力评估
Purpose & Capability
The name/description (multi-agent orchestration, tools, memory, delegation) match the SKILL.md content. All referenced capabilities (files, execute, memory, platform, browser, skills) are described and expected for a platform-runtime documentation skill.
Instruction Scope
The instructions are purely descriptive and show how to call platform tools (e.g., skills list/read, memory actions, execute curl to localhost). They assume the agent is running inside a SwarmClaw environment; those calls would be meaningless outside that runtime. The doc states credentials are injected as env vars and redacted — a policy claim that cannot be verified from this file but is relevant to trust decisions.
Install Mechanism
There is no install spec and no code files. This is instruction-only, so nothing is downloaded or written to disk by the skill itself.
Credentials
The skill declares no required environment variables or credentials. It documents that the platform makes configured credentials available as env vars (e.g., OPENAI_API_KEY), which is coherent for a runtime doc. It does not ask for unrelated or excessive secrets.
Persistence & Privilege
The skill does not request always:true and does not modify other skills or agent-wide configuration. It documents persistent memory and workspace conventions, which is appropriate for a runtime description.
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 swarmclaw - 安装完成后，直接呼叫该 Skill 的名称或使用
/swarmclaw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
v2.4.1
Fix display name casing
v2.4.0
Add proper ClawHub YAML frontmatter with metadata, privacy policy, data handling, version, author, homepage, and tags
v2.3.0
Add dreaming (idle-time memory consolidation) documentation
v2.2.0
- Added new "Goals" command group for hierarchical goal management.
- Skill description and documentation updated to include "goals" alongside agents, tasks, chats, chatrooms, schedules, etc.
- "Knowledge" command group documentation added, covering management of knowledge sources.
- Version bumped to 2.2.0.
v2.0.0
SwarmClaw Skill 2.0.0 — Major CLI/API Expansion
- Expanded skill capabilities to cover 40+ CLI command groups, including agents, tasks, chats, chatrooms, missions, schedules, memory, autonomy, connectors, wallets, extensions, and more.
- Updated documentation to reflect a full 1:1 mapping with the SwarmClaw REST API, supporting advanced filtering, bulk actions, and structured output.
- Added detailed command examples for agents, tasks, chatrooms, collaborative workflows, schedules, memory, missions, and global search.
- Introduced support for advanced output flags (`--json`, `--query`), new orchestration patterns, and comprehensive fleet management.
- Clarified and renamed core terminology (e.g., "sessions" are now "chats"), and described new advanced orchestration and agent automation capabilities.
v1.0.2
Remove remote install script recommendation; npm-only setup
v1.0.1
Add setup, multi-gateway fleet guidance, and cross-gateway status example
v1.0.0
Initial release — manage SwarmClaw agents, tasks, sessions from OpenClaw
元数据
常见问题
SwarmClaw 是什么？
AI agent runtime and multi-agent orchestration platform. Teaches agents how to use SwarmClaw's 6 primitive tools, persistent memory, dreaming, delegation, co... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 297 次。
如何安装 SwarmClaw？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install swarmclaw」即可一键安装，无需额外配置。
SwarmClaw 是免费的吗？
是的，SwarmClaw 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
SwarmClaw 支持哪些平台？
SwarmClaw 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 SwarmClaw？
由 Wayde（@waydelyle）开发并维护，当前版本 v2.4.1。
推荐 Skills