editTestCase MCP tool returns API_INTERNAL_ERROR on any call

Hi Testiny team,

We are integrating the Testiny MCP server (https://app.testiny.io/api/v1/mcp-server) into a Claude Code AI workflow and have run into a consistent issue with editTestCase.

Behavior:
Every call to editTestCase returns API_INTERNAL_ERROR, regardless of the payload. We tried editing only the automation custom field, only the title, and also sending a fuller update payload, but the result is always the same.

What works:
createTestCase, listTestCases, listTestCaseFolders, and listProjects all work correctly.

What does not work:
editTestCase fails on every attempt.

Could you confirm whether this is a known issue with the MCP endpoint? Is there a specific payload format required for editTestCase, or could the tool currently be broken server-side?

For now, we have fallen back to the REST API using PUT /api/v1/testcase/{id} with _etag, which works fine. So this is not a blocker, but using the MCP tool directly would be cleaner if it is expected to work.

Hello Evgeny,

I can’t directly reproduce your issue, editTestCase works fine in my test project and our test suites.

However, I noticed that our tool schema doesn’t require "project" or "project_id" in this call, while the correct schema would be to require either "project" or "project_id", so the model might omit it when sending the tool call, which results in API_INTERNAL_ERROR in my test.

If this is the issue you are running into, try adding "include the 'project' parameter with the project id as string" or something similar to your prompt. We will try to improve the schema and the error message for this case so the LLM doesn’t get confused.

if not, please try to post the the entire error JSON returned by the MCP server (from the Claude Code debug logs, with a timestamp) so we can try to diagnose it further

Regards,
Michael