The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill facilitates reading secrets, keys, and certificates from Azure Key Vault. If an attacker stores malicious instructions within these vault items, and the agent processes that content without sanitization, it could influence the agent's behavior.
Ingestion points: Data enters the agent's context through client.get_secret() and client.get_key() calls documented in SKILL.md.
Boundary markers: The instructions do not define delimiters or specific safety instructions to ignore embedded commands within retrieved data.
Capability inventory: The skill possesses significant capabilities including reading, writing, and deleting cryptographic materials and secrets within the Azure environment.
Sanitization: There is no evidence of validation or sanitization of the values retrieved from the Key Vault before they are used by the agent.
[SAFE]: All identified dependencies, including azure-keyvault-secrets, azure-keyvault-keys, azure-keyvault-certificates, and azure-identity, are official SDKs from Microsoft.
[SAFE]: No obfuscation, hardcoded credentials, persistence mechanisms, or unauthorized privilege escalation attempts were detected. The use of DefaultAzureCredential follows security best practices for the intended environment.

azure-keyvault-py