service-de-channel-consent-configure
Warn
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill performs shell-based operations using the
sfCLI tool to query and update records within the Salesforce environment. - [INDIRECT_PROMPT_INJECTION]: The skill accepts multiple inputs from the caller and interactive user prompts (Stage 3 and 4), including
{CHANNEL_ID},{ORG_ALIAS}, and various message strings like{OPT_IN_PROMPT}. - Ingestion points:
{CHANNEL_ID},{CONSENT_TYPE},{LANGUAGE},{ORG_ALIAS}, and user-defined prompt strings in Stage 3 and 4. - Boundary markers: None present; values are interpolated directly into command strings.
- Capability inventory: Execution of shell commands via
sf data queryandsf data update record. - Sanitization: No explicit sanitization or validation logic is provided to filter shell metacharacters from user inputs.
- [DYNAMIC_EXECUTION]: In Stage 5, the skill constructs shell commands by concatenating user-supplied strings into command templates. While the instructions mention a quoting rule for handling spaces, this pattern is susceptible to command injection if a user provides values containing shell-sensitive characters (such as backticks or single quotes) designed to break out of the intended command context.
Audit Metadata