Context Injection
An engineering method that embeds external information directly into the prompt sent to a large model as high-priority context. In contrast to letting the model retrieve information autonomously, it offers high determinism, controllable token usage and focused attention.
Timeline (last 90 days)
Context injection refers to the process in which, when calling the LLM, the agent concatenates specific text content into the model's input prompt, thereby influencing the model's output.
Context injection can update knowledge at any time without training, but consumes window budget on every call and is constrained by retrieval quality.
Cursor's @-Mentions are essentially context injection, directly embedding the content of the referenced object into the prompt sent to the model, existing as high-priority context.
All Facts (3)
Context injection refers to the process in which, when calling the LLM, the agent concatenates specific text content into the model's input prompt, thereby influencing the model's output.
50%UnverifiedContext injection can update knowledge at any time without training, but consumes window budget on every call and is constrained by retrieval quality.
50%UnverifiedCursor's @-Mentions are essentially context injection, directly embedding the content of the referenced object into the prompt sent to the model, existing as high-priority context.
50%