benchling-integration

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The skill includes examples that embed API keys and client secrets directly as string literals (e.g., ApiKeyAuth("your_api_key"), client_secret="your_client_secret"), which would require an LLM to output secret values verbatim if real credentials are substituted, creating an exfiltration risk.