The most novel thing in the Claude Code snapshot isn't a tool or a prompt. It's how aggressively the runtime is built around preserving prompt-cache identity across forks, compaction, and long-running sessions.

The most important file in the Claude Code snapshot isn't a prompt file or a tool definition. It's a state machine. Once you see that, a lot of agent design decisions start making sense.

After auditing a publicly posted Claude Code source snapshot and reading OpenAI's harness essays, I came away with one conclusion: the model is only the brain. The product is the runtime around it.