relational-database-mcp-cloudbase
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCREDENTIALS_UNSAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill defines workflows for the agent to query and process data from a relational database, which serves as an ingestion point for untrusted content that could contain hidden instructions intended to influence agent behavior.
- Ingestion points: The
queryMysqlDatabasetool is used to read SQL data into the agent context (SKILL.md). - Boundary markers: The instructions do not define specific delimiters or instructions for the agent to treat database results as untrusted data.
- Capability inventory: The skill provides high-privilege capabilities including database provisioning, destruction, schema mutation (
manageMysqlDatabase), and permission management (managePermissions). - Sanitization: No sanitization, filtering, or validation steps are prescribed for the data retrieved from the database before the agent processes it.
- [CREDENTIALS_UNSAFE]: The skill documents the
getConnectionInfoaction which is designed to expose raw database connection credentials for migration purposes. - Evidence: SKILL.md explicitly states that
action="getConnectionInfo"returns the raw connection/cluster payload and "may include credentials."
Audit Metadata