Schema validation error for `list_trades` tool missing type field for timestamp parameters

[kaligautier]

Description

When using the Massive MCP server (v0.6.0) with Google ADK and Gemini models, the list_trades tool causes a schema
validation error because several timestamp parameters are missing their type field in the schema definition.

Error Message

400 INVALID_ARGUMENT. {'error': {'code': 400, 'message': "Unable to submit request because list_trades
functionDeclaration parameters.timestamp_lt schema didn't specify the schema type field. Learn more:
https://cloud.google.com/vertex-ai/generative-ai/docs/multimodal/function-calling", 'status': 'INVALID_ARGUMENT'}}

Similar errors occur for other timestamp parameters:

• timestamp_gte
• timestamp_lte
• timestamp

Environment

• MCP Server Version: v0.6.0 (git+https://github.com/massive-com/mcp_massive@v0.6.0)
• Client: Google ADK (Agent Development Kit) with Gemini 2.0-flash
• Framework: Using McpToolset from google.adk.tools.mcp_tool
• Connection Type: STDIO

Steps to Reproduce

1. Set up Google ADK with Gemini model
2. Connect to Massive MCP server v0.6.0 via stdio:

from google.adk.tools.mcp_tool import McpToolset, StdioConnectionParams
from mcp.client.stdio import StdioServerParameters

server_params = StdioServerParameters(
    command="uvx",
    args=["--from", "git+https://github.com/massive-com/mcp_massive@v0.6.0", "mcp_massive"],
    env={"MASSIVE_API_KEY": api_key}
)
toolset = McpToolset(connection_params=StdioConnectionParams(server_params=server_params))
3. Create an agent with the MCP toolset
4. Try to run any query that might use list_trades

Expected Behavior

All tools should have properly defined schemas with type fields for all parameters, allowing Gemini to validate and
use the tools.