AI Workflow Test Input

What it's for

• Workflow orchestration canvas where the author tests an agent graph with a mock payload before publishing to production.
• Integration sandbox where corrupted payloads are injected to check that fallback nodes activate correctly.
• LLM pipeline debugging panel to inspect the raw input, the formatted output and token consumption.

For visual editing of graph nodes and connections is required instead of a payload test run, use AI Workflow Canvas. For periodic or cron runs need to be scheduled instead of ad-hoc one-off tests, use AI Workflow Schedule. For only atomic tool calls within an agent need to be inspected, use AI Tool Call Card.

Includes

• Payload preset selection
• Json payload editor
• Syntax validation
• Dry run execution
• Node trace inspection
• Output payload viewer
• Payload diff
• Execution metrics

Works with

• Radix UI primitives (the registry variant it ships in)
• React 19, Next.js App Router and Tailwind CSS 4, installed with the shadcn CLI
• Your theme tokens, in light and dark mode