DomainAgentHub
PassAudited by ClawScan on May 1, 2026.
Overview
This instruction-only skill is a task router with no code or required credentials, but it may delegate work to child agents and optional chat/external tools, so review those boundaries before use.
This skill appears coherent and benign for routing tasks to domain experts. Before installing, understand that complex tasks may be sent to child agents, optional Feishu/Gateway integrations may move messages between systems, and any suggested external CLI or API tool should be installed or connected only after you verify its source and permissions.
Findings (3)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
Complex requests may be delegated to specialized agents that have their own behavior and capabilities.
The skill explicitly instructs the agent to use a spawning/delegation tool for complex tasks. This matches the task-router purpose, but users should notice that task handling may move to other agents.
复杂任务、需要分析 | 用 `sessions_spawn` 分发给专业 Agent
Keep delegation visible to the user, show which agent will receive the task, and ask for confirmation before spawning agents for sensitive or high-impact work.
Task content and results could be shared with child agents or posted back into a configured group chat.
The workflow describes task and result flow through a local Gateway, child agents, and optionally a Feishu group chat. This is disclosed and aligned with the hub purpose, but the artifacts do not detail identity, permission, or data-boundary controls.
Gateway 监听 `ws://127.0.0.1:18789` ... 飞书群聊 ← 用户发送任务 ... Agent 执行完成 → MaxClaw 转发结果到群聊
Verify which chats and agents are connected, avoid sending sensitive data unless those channels are trusted, and confirm where results will be posted.
Following the recommended setup commands would install third-party tools globally on the user’s machine.
The skill gives user-directed examples for installing external global npm CLI tools. These examples are not automatic execution, but users should still verify package provenance and versions.
安装: npm install -g @anthropic-ai/claude-code ... 安装: npm install -g opencode
Install external tools only from trusted sources, review their permissions, and prefer pinned or official installation instructions.
