Expired50% confidenceEventExact time
A Bilibili UP loader shared an experiment based on DeepSeek Harness (DSH): having one agent supervise, debug, and evolve another agent software.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
9/15/2026
First Seen
This claim has expired. It is kept for the record and no longer describes the current state.
Valid until: 9/29/2026(expired)
Sources
Related Entities
Related Claims
UnverifiedDeepSeek Harness 的 Agent 在开发任务前会先给出交互式计划并创建可追踪的任务清单，采用「先规划、后执行、可追踪」的工作流72% similarUnverifiedIn LangChain's hosted Deep Agents, the model, prompts, tools, and sub-agents are determined by the developer, while persistence, memory, skill loading, code execution sandboxing, and deployment are handled by the platform.70% similarUnverifiedAgent Skill通过System Prompt工程、输出Schema约束和Few-shot示例注入来实现对LLM能力的专业化封装70% similarUnverifiedDeep Agent possesses a closed-loop 'perception-planning-execution' capability, and can not only generate code but also read project file trees, understand module dependencies, invoke terminal commands, parse compilation errors, and autonomously fix problems.69% similarUnverified科研自动化 Agent 通常基于 LLM 的工具调用与任务规划能力，结合代码执行沙箱、文件系统访问、学术检索及 GitHub/Hugging Face API 等外部工具，形成感知-规划-执行-反馈的闭环68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/923888API
curl https://kongchang.com/api/v1/knowledge/claims/923888MCP
get_claim(id=923888)