← 返回 Skills 市场
Multi Agent Workflow
作者
dxie48892-jpg
· GitHub ↗
· v1.0.3
· MIT-0
316
总下载
0
收藏
5
当前安装
4
版本数
在 OpenClaw 中安装
/install multi-agent-workflow
功能描述
多Agent协作工作流 v1.0.3。让多个AI Agent组成团队,协同完成复杂任务。支持任务拆分、角色分配、结果汇总、任务管理。
使用说明 (SKILL.md)
Multi Agent Workflow - 多Agent协作工作流
快速开始
cd scripts
python workflow.py --help
目录结构
multi-agent-workflow/
├── SKILL.md
├── README.md
├── scripts/
│ └── workflow.py
工作流程
- 接收任务
- 拆分任务
- 分配角色
- 并行执行
- 汇总结果
更新日志
v1.0.3 (2026-03-22)
- 修复:Windows控制台UTF-8编码输出问题
v1.0.2 (2026-03-21)
- 新增:TaskManager任务管理器
- 新增:split_task任务拆分
- 新增:自动角色分配
安全使用建议
This skill appears to be a simple, coherent task-splitting and role-assignment helper implemented as a local Python script. It does not access the network, request credentials, or perform installs. Before installing or running: 1) Inspect the script (you already have it) — it's small and readable. 2) Note that claims about 'parallel execution' and 'sessions_spawn' are not implemented; if you expect agent orchestration, additional integration code is required. 3) If you will run it on sensitive systems, run it in an isolated environment (e.g., container/VM) as a general best practice for any third‑party script. Overall risk is low, but only run it if you understand how you'll integrate it with your agent framework.
功能分析
Type: OpenClaw Skill
Name: multi-agent-workflow
Version: 1.0.3
The skill bundle provides a legitimate framework for managing multi-agent workflows by categorizing tasks and assigning roles. The core logic in `scripts/workflow.py` uses simple keyword matching and task management structures to output JSON configurations, with no evidence of data exfiltration, unauthorized execution, or malicious prompt injection.
能力评估
Purpose & Capability
Name/description match the delivered artifacts: SKILL.md, README, and scripts/workflow.py implement task splitting, role assignment, a TaskManager, and result aggregation. The 'model': 'iflow' fields are just descriptive strings and do not trigger network or credential use.
Instruction Scope
Runtime instructions are limited to running the included Python CLI. The SKILL.md and code do not read system files or environment variables. Minor mismatch: meta/README/description refer to '并行执行' (parallel execution) and '可用于调用OpenClaw的sessions_spawn执行', but the provided script does not implement parallel execution or any auto-spawning of agent sessions — it only prints JSON and manages in-memory tasks.
Install Mechanism
No install spec is provided and no external packages are fetched; the skill is instruction + a small local Python script only, which is low risk from an installation standpoint.
Credentials
The skill requests no environment variables, credentials, or config paths. The code does not access secrets or external credentials.
Persistence & Privilege
Flags show normal privileges (always: false, user-invocable: true). The skill does not persist or modify other skills or system configuration.
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 multi-agent-workflow - 安装完成后,直接呼叫该 Skill 的名称或使用
/multi-agent-workflow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
v1.0.3
v1.0.3: 修复Windows控制台UTF-8编码输出问题
v1.0.2
v1.0.2: 新增TaskManager任务管理器、split_task任务拆分、自动角色分配
v1.0.1
修复:补充Agent调度代码
v1.0.0
多Agent协作工作流 - 任务拆分、角色分配、结果汇总
元数据
常见问题
Multi Agent Workflow 是什么?
多Agent协作工作流 v1.0.3。让多个AI Agent组成团队,协同完成复杂任务。支持任务拆分、角色分配、结果汇总、任务管理。 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 316 次。
如何安装 Multi Agent Workflow?
在 OpenClaw 或 Claude Code 对话框中运行命令「/install multi-agent-workflow」即可一键安装,无需额外配置。
Multi Agent Workflow 是免费的吗?
是的,Multi Agent Workflow 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。
Multi Agent Workflow 支持哪些平台?
Multi Agent Workflow 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 Multi Agent Workflow?
由 dxie48892-jpg(@dxie48892-jpg)开发并维护,当前版本 v1.0.3。
推荐 Skills