Cursor agents can show where their context window went
By Steven Van ·
A new interactive report breaks down token usage by system prompt, tools and rules to help debug bloated agent runs.
Cursor agents can now break down their context window as an interactive report inside a canvas, showing how many tokens go to the system prompt, tools, rules and other components.
For anyone hitting context limits or trying to figure out why an agent run is bloated, that's a direct way to see where the tokens are going rather than guessing. Read the announcement for the full details, or check it out in Cursor.
Cursor
AI code editor built on VS Code — agents that edit across your codebase, run in the cloud, review pull requests, and now host your repos.
View Cursor →