Unverified50% confidenceFactExact time
LLM本身是无状态的，每次对话独立，跨会话记忆需依靠外部文件实现
1
Sources
50%
Confidence
Long-term
Relevance
7/12/2026
First Seen
Sources
Andrew Ng × Anthropic: A Hands-On Claude Code Tutorial and Core Methodology
bilibiliAI从零开始学7/9/2026
Related Claims
UnverifiedLLMs are essentially stateless; after each conversation ends, the model itself retains no information.78% similarUnverifiedThe core capability of LLMs is confined to the 'text input → text output' framework, making them unable to proactively read project files, execute system commands, or access external tools.64% similarUnverifiedSingle-session instructions are only valid during the current session and expire upon closing, whereas global custom instructions, written once, take effect permanently and apply across projects.61% similarUnverifiedLLM 本身不具备直接操作文件系统、调用终端或感知代码库拓扑结构的能力，这些工程能力必须由 IDE 层封装实现61% similarUnverifiedFengxing Story does not include a built-in LLM and relies entirely on external LLM APIs to function.58% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/489953API
curl https://kongchang.com/api/v1/knowledge/claims/489953MCP
get_claim(id=489953)