Malformed_Tool_Response_Handling

Scenarios testing how the agent handles malicious or malformed structured data returned by tools: deeply nested JSON, type confusion, unexpected or duplicate fields.

▶ Try the tools (generate tools first)

Generator controls

|/api/mcp/srv-malformed-tool-response-handling

MCP enable + token are only for calling from an external client — the playground below needs neither.

Tools (0)

None yet — hit Generate tools.

Scenarios (0)

None yet — create one after generating tools.