The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill generates XML configuration files (ssma-variables.xml and ssma-servers.xml) on the local filesystem that contain plaintext database passwords for both Oracle and SQL Server environments.
[COMMAND_EXECUTION]: The skill invokes a local system executable (SSMAforOracleConsole.exe) via PowerShell using dynamic parameters. While this is the primary function of the skill, it utilizes user-provided paths and configuration strings in a shell execution context.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a vulnerability surface for indirect prompt injection as it processes database schema data.
Ingestion points: Data is ingested through Oracle schema names, service names, and resulting assessment reports read back into the agent context.
Boundary markers: There are no delimiters or specific instructions to ignore embedded commands when the agent processes schema metadata or report outputs.
Capability inventory: The skill has the capability to write files (XML configs), create directories, and execute shell commands.
Sanitization: The instructions do not specify any sanitization or validation for database identifiers before they are interpolated into configuration files or shell commands.

ssma-console