sql-server-dba-dev-expert

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt includes a hardcoded plaintext password ('SecureP@ssw0rd#123!') in the encryption key example, which is a secret value that the LLM would need to reproduce verbatim if asked to output or reuse the example—this is high-risk secret handling.

HIGH W008: Secret detected in skill content (API keys, tokens, passwords).

• Secret detected (high risk: 1.00). I scanned the skill content for literal, high-entropy credential material. I found a clear password literal in the "Sensitive Data Encryption" section:

CREATE MASTER KEY ENCRYPTION BY PASSWORD = 'SecureP@ssw0rd#123!';

This is a non-placeholder, non-trivial password used to encrypt the database master key — a real secret by definition (it can unlock cryptographic material). It is not a generic placeholder or obvious example password.

Other strings (backup file paths, DB names, role names, sample SQL, URLs, and simple example values) are either environment/config names, file paths, or low-entropy/example values and do not meet the "secret" criteria, so I ignored them.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 0.90). The prompt instructs the agent to execute many state-changing and security-affecting SQL/server operations (CREATE/ALTER objects, INSERT/UPDATE/DELETE, CREATE MASTER KEY with plaintext password, ALTER LOGIN sa DISABLE, BACKUP/RESTORE WITH REPLACE, DBCC REPAIR_ALLOW_DATA_LOSS, etc.), which directly modify database/server configuration and can be destructive, so it pushes the agent to change and potentially compromise the host system's state.