Studio Workbench
Try the API without writing code — type symbols, run, and see the exact JSON your agent would receive. Guard checks any LLM sentence against real data.
Loading your account…
Checking account…
Paste LLM output to validate against last session (Agent Pro / Guard).
Create a session in Context tab first.
Preview how context fits in a system prompt (no LLM call).
Run context first.
Save user-selected symbols for a non-personalized watchlist brief. The brief does not use your financial situation or assess suitability. Email delivery ~07:30 Paris (Developer, Agent Pro, Research).
—