AI Agent Hub
Back to plugins
⚙️

dsh-openclaw-enterprise-team

Workflow Updated 2026.08.18

Run the following command in DeepSeek Harness:

dsh plugin install youwenkui/dsh-openclaw-enterprise-team

Paste the following prompt into your AI chat to install this plugin:

Run dsh plugin install youwenkui/dsh-openclaw-enterprise-team inside DeepSeek Harness to install this plugin; the source repository is at https://github.com/youwenkui/dsh-openclaw-enterprise-team .

About this plugin

DeepSeek Harness is a conversational interface for developers, yet many enterprises already run an OpenClaw AI team as the authoritative system for agents, skills, tools, memory, and execution state. This plugin bridges the two without duplicating anything: DSH becomes the natural-language entry point, while OpenClaw remains the single source of truth for every session, approval, and execution record.

Four purpose-built tools cover the full workflow. openclaw_team_health verifies Gateway availability through a minimal, redacted projection. openclaw_team_list discovers the agents currently configured. openclaw_team_start hands a task to a designated lead agent for execution. openclaw_team_continue steers the same session for corrections, added requirements, or approval responses. Every task returns an OpenClaw-native sessionKey, so no second task identifier is ever invented or persisted inside DSH. On the security side, the package stores no keys or tokens, executes commands via execFile rather than a shell, and redacts sensitive fields across health responses, agent lists, and result outputs.

This is designed for teams that already operate a healthy OpenClaw Gateway with at least one agent configured and want to direct their enterprise AI workflow through DSH's natural-language interface while keeping OpenClaw authoritative over permissions, approvals, and data. The plugin is currently in experimental alpha and requires DeepSeek Harness 0.1.0-rc.7 alongside Node.js 22 or newer.

Use Cases

  • Direct an OpenClaw lead agent to research via natural language, then refine the same session using the returned sessionKey
  • Verify OpenClaw Gateway health, discover configured agents, and launch a task—all from within the DSH conversation
  • Respond to approvals or add requirements to a live OpenClaw session without leaving DSH or dropping to the CLI

Best For

  • Technical teams already running an OpenClaw Gateway with enterprise AI agents
  • Developers who prefer driving enterprise workflows through DSH conversational UI over raw CLI
  • Architects who need OpenClaw as the sole source of truth without duplicating agents or sessions inside DSH