connect-tool
Pass
Audited by Gen Agent Trust Hub on Nov 2, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands using tools such as
psql,sqlcmd, andmdb-export. These commands are used to validate database connectivity and, crucially, to verify that read-only accounts cannot perform write operations (e.g., attempting to create a table or insert values and checking for permission denied errors). - [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources during the research phase, creating a potential surface for indirect prompt injection.
- Ingestion points: The agent is directed to research vendor documentation, official developer portals, and the
registry.modelcontextprotocol.ioregistry. - Boundary markers: The instructions do not specify the use of delimiters or specific instruction-ignore wrappers for the content retrieved from external URLs.
- Capability inventory: The agent has the capability to write to the local filesystem (under
01-TOOLS/) and execute database client utilities in the shell. - Sanitization: No explicit sanitization or filtering of the researched technical documentation is mentioned before the agent uses it to formulate connection strategies.
Audit Metadata