Unverified50% confidenceFactExact time
所有LLM本质上都是高级的自动补全，模型本身无法编辑文件、操作数据库、连接外部服务或上网，只能生成文本
1
Sources
50%
Confidence
Long-term
Relevance
7/10/2026
First Seen
Sources
Related Claims
UnverifiedThe 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.65% 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63% similarUnverifiedAWS does not use customer code or prompts to train models, nor does it share this information with model providers, and telemetry data originally sent to Anthropic through the Bedrock runtime is disabled by default.60% similarUnverifiedLLM的代码不只是源代码文件，还包括提示词、RAG配置、模型版本，这些传统CI/CD系统并不感知59% similarUnverifiedLLMs are essentially stateless; after each conversation ends, the model itself retains no information.58% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/465371API
curl https://kongchang.com/api/v1/knowledge/claims/465371MCP
get_claim(id=465371)