DeepSeek Super-agent design and multi-agent engineering practices
Author Cang He summarizes the DeepSeek Harness team’s Super-agent design, including parallel Workers with independent contexts, direct delivery of follow-up messages through subscription channels, database-backed session state, and an external Watchdog for monitoring execution and exceptions. He then describes his own LobeHub workflows for delegation, recurring inspections, local Codex and Claude Code connections, code review, and acceptance evidence. The account also documents unresolved document-access and review-configuration issues and says that the initial review was not independently verified.
The account presents concrete patterns for preserving task context across multiple agents, carrying work from review into implementation and verification, and retaining evidence for recurring tasks. It is an author’s engineering account based on the DeepSeek design, not an independently confirmed DeepSeek product announcement.
这是苍何的第 617 篇原创!
大家好,我是苍何。
最近看了 DeepSeek Harness 团队的《Super-agent 设计探索,迈向更自主的 Agent 系统》,感触还是很深的。
过去一年,Agent 的单体能力在大幅增强,但很多还困在单次的对话里,切换对话,还要主动记忆上下文,而同时进行的对话 Session 一多,人的注意力就会跟不上了。
最近很火的 Personal Agent 在个人助手场景下可以持续观察信息流并协助处理事务,而且具备全局感知的上下文,是 Super-agent 比较好的一种形态。
Some source content is omitted. Continue reading on the original site.
#DeepSeek#DeepSeek Harness#Super-agent#LobeHub#Codex#Claude Code