Context Explorer
A language model has no memory. On every turn, the software around it, called the harness, sends everything again: the instructions, the tool definitions, the conversation so far and anything the model has read. These are scripted examples of agents doing everyday work. After each reply, press Show context to see exactly what the model received, how much room each part takes and what is new. Pick an example to start.
Example model
Token counts are approximate. Where a file or page is long, the text shown is an excerpt and the count is for the full block. Real harnesses add more tool definitions and formatting than shown here.