← Back to Skills Marketplace
fish1981bimmer

openclaw-collab

by fish1981bimmer · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ⚠ suspicious
66
Downloads
0
Stars
0
Active Installs
2
Versions
Install in OpenClaw
/install openclaw-collab
Description
Hermes 与本地 OpenClaw 协同工作 — 模型互调、记忆共享、任务分工
Usage Guidance
这是一个说明型协同指南,功能与描述基本一致,但在安装元数据中没有声明的凭证与能力出现在运行说明中。安装前请: - 检查并审计 ~/.hermes/scripts/openclaw-bridge.py 和 hermes-to-openclaw.py 的源码,确认它们不会无意上传或泄露文件/凭证。 - 确认 ~/.openclaw/workspace/ 中没有包含敏感凭证或私钥(或将这些文件移动到受限位置),因为指南允许 Hermes 读写该目录。 - 如果你会用到 openclaw agent --local 或 clawhub publish,明确哪些 API key/凭证将被使用,在哪个进程/环境中可见,并在元数据中记录这些需求或限制访问范围。 - 在首次运行时,以非特权账户和受限示例数据进行测试,观察 hermes-to-openclaw.py 是否向本地网关或远程主机发出意外网络请求。 如果开发者能提供这两个脚本的源码或补充声明(需要哪些环境变量、clawhub 发布是否需要凭证并如何提供),可以将评估升级为“benign”。
Capability Analysis
Type: OpenClaw Skill Name: openclaw-collab Version: 1.0.0 The skill bundle contains metadata and documentation (SKILL.md) defining a collaboration framework between two AI agents, Hermes and OpenClaw. It provides instructions for task delegation, memory sharing via a local workspace (~/.openclaw/workspace/), and model interaction using local bridge scripts (~/.hermes/scripts/openclaw-bridge.py). No malicious logic, data exfiltration, or harmful prompt injections were identified; the content is strictly focused on operational coordination and workflow optimization between the agents.
Capability Tags
cryptorequires-sensitive-credentials
Capability Assessment
Purpose & Capability
技能名与描述(Hermes 与本地 OpenClaw 协同)与 SKILL.md 中的内容一致:指南说明如何通过 bridge 和唤醒脚本进行模型互调、如何读写 ~/.openclaw/workspace/ 的 memory/、以及如何分工和发布到 ClawHub。没有看见与目的不相关的依赖或不相干的二进制请求。
Instruction Scope
SKILL.md 明确指示运行本地脚本(~/.hermes/scripts/openclaw-bridge.py、hermes-to-openclaw.py)、使用 openclaw CLI、以及读取/写入 ~/.openclaw/workspace/ 下的文件。整体在协同目的范围内,但这些操作会访问用户家目录下的文件并可触发本地/网络发布流程(clawhub publish),应注意这些文件可能包含敏感信息或凭证。说明还提到可使用通用 read_file/write_file/patch 操作,这赋予了代理对 workspace 中文件的广泛读写权限。
Install Mechanism
这是一个 instruction-only 的技能,没有安装步骤或第三方下载;这降低了写磁盘/执行远程代码的直接风险。
Credentials
技能元数据不声明任何环境变量或凭证,但文档中多处暗示或显式提到需要凭证或本地 API key:例如“openclaw agent --local 需要 shell 中有 API key”、以及将 skill 发布到 clawhub.ai(clawhub CLI 通常需要身份认证)。这构成不一致:说明要求凭证/网络发布能力,但清单未要求任何 env/凭证,用户可能未被告知需要提供哪些密钥或将被使用。
Persistence & Privilege
技能没有设置 always:true,也没有修改其他技能或系统范围配置的说明。唯一的持久影响是指示读/写 ~/.openclaw/workspace/(这是其功能所需的常规范围),但技能本身不会被强制常驻。
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install openclaw-collab
  3. After installation, invoke the skill by name or use /openclaw-collab
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Hermes与OpenClaw协同工作 - 模型互调、记忆共享、任务分工
v1.1.0
v1.1.0 更新:添加hermes-to-openclaw.py工具说明,支持直接唤醒OpenClaw处理消息,无需轮询;更新常见错误处理;完善文档
Metadata
Slug openclaw-collab
Version 1.0.0
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 2
Frequently Asked Questions

What is openclaw-collab?

Hermes 与本地 OpenClaw 协同工作 — 模型互调、记忆共享、任务分工. It is an AI Agent Skill for Claude Code / OpenClaw, with 66 downloads so far.

How do I install openclaw-collab?

Run "/install openclaw-collab" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is openclaw-collab free?

Yes, openclaw-collab is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does openclaw-collab support?

openclaw-collab is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created openclaw-collab?

It is built and maintained by fish1981bimmer (@fish1981bimmer); the current version is v1.0.0.

💬 Comments