AGENT WORKSPACE
Coordinate the solve.
Launch focused teams, follow their evidence, and keep every candidate flag accountable.
0runs
0active
...model
Per agenthosted browsers
NEW RUN
Give the team a challenge
HISTORY
Runs
NO RUN SELECTED
A place for the evidence.
Launch a team or choose a run to see agent work, tool output, and verified flags.
Agent team
Challenge files
Flag verification
Evidence log
0 eventsDIRECT MODEL
Ask without a run.
This conversation uses the configured model directly. Agent teams provide tools, browser access, and recorded evidence.