← 返回 Skills 市场
104
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install eo-ability-multi-expert
功能描述
多专家编排能力(Multi-Expert Orchestrator),协调多个专家并行/串行工作,结果自动汇流
使用说明 (SKILL.md)
eo-ability-multi-expert
多专家编排能力 (Multi-Expert Orchestrator) - 协调多个专家并行/串行工作
一句话介绍
EO核心编排能力,多专家并行启动、结果自动汇流、依赖管理、冲突解决。
核心功能
- 多专家并行: 多个专家同时启动
- 结果汇流: 专家输出自动合并
- 依赖管理: 专家执行顺序控制
- 冲突解决: 多专家结果冲突处理
- Checkpoint: 每N个专家后自动Checkpoint
使用方法
# 启动多专家协作
/dream "开发博客系统"
// 或通过 Team Manager
创建团队: 博客开发团队
添加专家: Architect, Backend, Frontend, QA
启动协作: parallel
与EO插件的协同
- 被所有 eo-workflow-* 调用
- 是 EO 的核心差异化能力
独立运行模式(有EO vs 无EO)
| 模式 | 能力 |
|---|---|
| 有EO插件 | 真实多Agent并行、sessions API调用、结果汇流 |
| 无插件(基础) | LLM模拟多专家、串行执行 |
执行流程
用户: /dream "开发博客系统"
│
▼
┌─────────────────────────────────────┐
│ Multi-Expert Orchestrator │
│ 解析任务 → 分解 → 调度 │
└─────────────────┬───────────────────┘
│
┌─────────────┼─────────────┐
▼ ▼ ▼
┌────────┐ ┌────────┐ ┌────────┐
│Architect│ │Backend │ │Frontend│
│ Agent │ │ Agent │ │ Agent │
└────┬────┘ └────┬────┘ └────┬────┘
│ │ │
└────────────┼────────────┘
▼
┌──────────────┐
│ Result Merge │
│ + Checkpoint │
└──────────────┘
Interface
Input
interface MultiExpertInput {
experts: ExpertRequest[]
strategy: 'parallel' | 'sequential' | 'pipeline'
checkpointAfter?: number // 每 N 个专家后 Checkpoint
timeout?: number // 超时时间(ms)
}
Output
interface MultiExpertOutput {
results: ExpertResult[]
checkpoints: CheckpointResult[]
summary: string
duration: number
}
🦞⚙️ 钢铁龙虾军团
安全使用建议
This skill is instruction-only and internally consistent, but before installing or enabling it: 1) Verify the skill's source and trustworthiness (homepage/source unknown). 2) If you pair it with an actual EO plugin or eo-workflow-* integrations, inspect those implementations—they may request network access or credentials not present here. 3) Watch for any runtime prompts that ask for API keys or system credentials and avoid granting unrelated secrets. 4) Prefer least-privilege for any EO/session integrations and enable auditing/logging so you can review what agents actually do.
功能分析
Type: OpenClaw Skill
Name: eo-ability-multi-expert
Version: 1.0.0
The skill bundle contains only metadata and documentation (SKILL.md) describing a multi-agent orchestration framework. There is no executable code, no installation scripts, and no instructions that suggest data exfiltration, unauthorized access, or malicious prompt injection. The content is entirely aligned with its stated purpose of coordinating parallel agent workflows.
能力评估
Purpose & Capability
The name/description (multi-expert orchestrator) align with the instructions: splitting tasks, scheduling experts, merging results, checkpointing. The skill declares no extra binaries, env vars, or config paths, which is proportionate to an instruction-only orchestrator.
Instruction Scope
SKILL.md contains orchestration logic and usage examples (/dream, Team Manager). It does not instruct the agent to read local files, system paths, or unrelated environment variables, nor does it instruct exfiltration or calls to unexpected external endpoints. It does reference EO sessions API and eo-workflow-* callers as integration points, which is expected for an orchestrator.
Install Mechanism
No install spec and no code files are present (instruction-only). This minimizes on-disk execution and supply-chain risk.
Credentials
The skill requests no environment variables or credentials. This is proportionate to the stated purpose. Note: actual integrations with an EO plugin or sessions API (mentioned in the doc) could require credentials in a concrete implementation—those are not present here.
Persistence & Privilege
always is false and default autonomous invocation is allowed (normal). The skill does not request permanent presence or claim to modify other skills or system-wide settings.
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 eo-ability-multi-expert - 安装完成后,直接呼叫该 Skill 的名称或使用
/eo-ability-multi-expert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
v1.0.0
Initial release of eo-ability-multi-expert — Multi-Expert Orchestrator capability.
- Supports orchestrating multiple experts in parallel or sequential execution.
- Automatically merges expert outputs and manages result dependencies.
- Handles expert result conflicts and supports checkpointing after configurable intervals.
- Can run independently or as a core EO plugin, adapting to presence or absence of EO.
- Exposes an interface for specifying experts, execution strategy, checkpoints, and timeouts.
元数据
常见问题
Eo Ability Multi Expert 是什么?
多专家编排能力(Multi-Expert Orchestrator),协调多个专家并行/串行工作,结果自动汇流. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 104 次。
如何安装 Eo Ability Multi Expert?
在 OpenClaw 或 Claude Code 对话框中运行命令「/install eo-ability-multi-expert」即可一键安装,无需额外配置。
Eo Ability Multi Expert 是免费的吗?
是的,Eo Ability Multi Expert 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。
Eo Ability Multi Expert 支持哪些平台?
Eo Ability Multi Expert 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 Eo Ability Multi Expert?
由 467718584(@467718584)开发并维护,当前版本 v1.0.0。
推荐 Skills