Return input validation errors as Tool Execution Errors

[jaylonmcshan19-x]

PCI review checklist

When tool input validation fails, errors were returned as Protocol Errors (JSON-RPC ). These errors were caught at the MCP Client level and never reach the LLM's context window. which causes:

• The Model retries sort of blindly with the same invalid inputs
• Repeated failures without any feedback loop
• leads to a less desired user experience due to there needing manual intervention

This PR creates utils.ToolError() and utils.ToolErrorf() helpers that return Tool Execution Errors (isError: true) instead of Protocol Errors (Existing flow). Tool Execution Errors are included in the JSON-RPC response body, making them visible to the LLM.

Example flow:

return nil, utils.LogAndReturnError(logger, "missing input", err)
// → Protocol Error → The Model never sees this, which will result in retries, and repeated failures

return utils.ToolError(logger, "missing input", err)
// → Tool Execution Error → Model can now sees this and can self-correct

• I have documented a clear reason for, and description of, the change I am making.

• If applicable, I've documented a plan to revert these changes if they require more than reverting the pull request.

• If applicable, I've documented the impact of any changes to security controls.

  Examples of changes to security controls include using new access control methods, adding or removing logging pipelines, etc.

[github-actions]

Changelog Warning

Please add a changelog entry to CHANGELOG.md for this change. If you believe this change does not need a changelog entry, please add the 'no-changelog-needed' label.