DeepSeek Harness Team Releases Super-agent Architecture

Related topics top start Related topics top end Futures promotion begin Futures promotion end Miaopai begin Miaopai end Video player start Video player end Market chart begin Market chart end

On the evening of October 31, the DeepSeek Harness team published a technical article on its official WeChat public account titled "Super-agent Design Exploration: Toward a More Autonomous Agent System."

DeepSeek stated that improvements in model capabilities have given rise to new application forms, with Personal Agents beginning to emerge in concentrated numbers. However, existing mainstream Agent designs still struggle to support this model at the scale of multitasking and team collaboration, and the upper limit on multitasking depends on human attention.

To break through these limitations, it is necessary to break the binding between the Agent and a single manual session, turning it into a continuous entity that can maintain context across sessions, advance multiple complex tasks simultaneously without interfering with one another. DeepSeek calls this Agent form and its system architecture Super-agent.

Through approaches such as a dual-layer architecture, event-driven design, asynchronous multi-channel interaction, separation of storage and compute, and an external Watchdog, the DeepSeek team overcame these limitations. At the underlying level, the system takes over cumbersome session management and context synchronization, allowing humans to move further up from specific execution details to higher-level decision-making and acceptance.

Specifically, Super-agent has the following four basic characteristics:

Continuous: Users can interact with it directly in familiar work interfaces such as IM chats and GitHub comments, and it feels as if there is always only one Agent, with no need to think about "whether Session X's context contains information Y." For example, if a user first proposes adopting Plan A on IM and later changes to Plan B in a GitHub Pull Request comment, it can understand and adjust the direction of execution.

Efficient: It responds quickly to interactions while completing multiple complex tasks in parallel and with high quality. It can accept the next instruction without waiting for the previous command or task to be completed.

Proactive: It does not need to wait for users to issue instructions one by one; it can continuously observe environmental events and proactively advance matters worth handling, while staying quiet when no intervention is needed.

Persistent: It is always online, connects directly to the user's real work environment, and automatically recovers from anomalies encountered during long-term operation.

The article mentioned that at present, an internal Super-agent instance built on this approach is already running continuously, participating in daily issue reproduction and fixes, code review responses, and iteration of DeepSeek Harness itself. When team members report software issues in group chats, even if no one actively @s the AI, it can identify potential defects and attempt in the background to reproduce the issue and find a fix.

DeepSeek believes that in this form, the Agent is no longer just a dialogue window passively waiting for instructions, but has evolved into an autonomous collaborator that is continuously online in the digital environment: for individuals, it can be a personal assistant that helps handle daily affairs across devices; for teams, it integrates into R&D workflows, taking on and advancing specific tasks. As model capabilities and underlying system architecture evolve together, this more autonomous way of working is expected to become the norm in human-computer collaboration. article_adlist[

Source: DeepSeek Harness team WeChat public account]article_adlistarticle_adlist[

]article_adlist news_keyword_pub,stock, iframe below the main text iframe below the main text General statement General statement QR code at the end start QR code at the end start

QR code at the end start QR code at the end start Editor name and work code

Editor name and work code