Claude MD
A Markdown file introduced by Anthropic that is injected into the model's context at the start of a session. It is used to store information such as environment instructions, organizational conventions, and personal preferences, and is a foundational practice of context engineering.
Timeline (last 90 days)
Claude MD is a Markdown file injected into the model's context at the start of a session, containing information such as environment hints, organizational norms, and personal preferences.
Claude MD is injected wholesale into the model's context at the start of a session, so it crowds out the space available for the model to handle actual tasks—the larger the file, the less space is left for reasoning and tool output.
The gradual rollout path Lamis outlined is: first use simple means such as Claude MD and skills; when the number of agents grows, run cycles lengthen, and collaboration becomes more complex, introduce production-grade mechanisms such as versioning, hashing, and tiered permissions; finally consider adding out-of-session processes like Dreaming.
All Facts (5)
Claude MD文件是Claude Code的记忆系统，用于告诉Claude Code关于团队目标、代码架构、代码库中的陷阱和最佳实践
90%UnverifiedClaude MD文件的内容会被注入到每次对话的上下文窗口最前端，属于系统提示词的持久化形式
80%UnverifiedClaude MD is a Markdown file injected into the model's context at the start of a session, containing information such as environment hints, organizational norms, and personal preferences.
50%UnverifiedClaude MD is injected wholesale into the model's context at the start of a session, so it crowds out the space available for the model to handle actual tasks—the larger the file, the less space is left for reasoning and tool output.
50%UnverifiedThe gradual rollout path Lamis outlined is: first use simple means such as Claude MD and skills; when the number of agents grows, run cycles lengthen, and collaboration becomes more complex, introduce production-grade mechanisms such as versioning, hashing, and tiered permissions; finally consider adding out-of-session processes like Dreaming.
50%