"ai slop" is no longer the fault of the model, but the fault of the user. in black box systems like claude code, context is the single input we can control - so how do we optimize it? ## what is context? context refers to everything you provide an llm when you send it a message. this includes the prompt itself, along with all of the surrounding information; system prompts, metadata, your previous messages, the llm's thinking, tool calls, and responses - everything. llms have limited…