The Kiota Python generator improperly handles enum value descriptions from the x-ms-enum.values[].description field in OpenAPI specifications. Specifically, newline characters are not sanitized, allowing attacker-controlled input to break out of comment contexts and inject executable Python code at the module scope. This code executes upon import of the generated Python module. The vulnerability requires user interaction in generating and importing code from a malicious specification, commonly occurring in development or CI workflows. The root cause is twofold: a missing cleanup step for enum descriptions and incomplete sanitization of newline characters in Python code generation. Exploitation vectors include consuming malicious OpenAPI specs from attacker-controlled URLs or supply-chain scenarios. The vendor has released patches in Kiota versions 1.29.1, 1.32.0, and later to address this issue.

Successful exploitation allows arbitrary Python code execution at import time of generated client modules. This can lead to credential theft, exfiltration of environment variables and secrets, source code disclosure, and persistent backdoors via regenerated client code. The impact primarily affects developers and CI/CD pipelines that generate and import Python SDKs from untrusted or malicious OpenAPI specifications.

Upgrade Kiota to version 1.29.1, 1.32.0, or later to apply the official fix. Regenerate any previously generated Python clients to remove vulnerable code. If immediate upgrade is not possible, only generate clients from trusted, internally controlled OpenAPI specifications, disable or gate automatic generation from remote specs in CI/CD pipelines, validate and sanitize x-ms-enum descriptions before generation, and review generated Python files for suspicious top-level code near enum declarations. Run generation and validation in isolated environments without access to production secrets. Confirm after upgrade that malicious multi-line enum descriptions no longer produce executable top-level output.