azuresql-db-auth

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The skill contains explicit password literals and connection-string examples (e.g., CREATE LOGIN ... WITH PASSWORD = 'An0ther_Str0ng_Passw0rd' and Password=... in examples) which would require the LLM to emit secret-like values verbatim if it reproduces those snippets, creating an exfiltration risk despite recommending secure stores elsewhere.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill declares a required container image that would be fetched and executed at runtime (sqldbpreview-dpgaeqhmgphzd4bk.azurecr.io/azure-sql/db-dev:latest), so this external artifact can run remote code when the skill is used.