commit
Pass
Audited by Gen Agent Trust Hub on Mar 14, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute a local bash script located at
/mnt/skills/user/commit/scripts/validate-commit.shto perform message validation. - [PROMPT_INJECTION]: Potential for indirect prompt injection and shell command injection via untrusted data ingestion.
- Ingestion points: The
messageparameter in the validation command withinSKILL.mdaccepts arbitrary text, which may be derived from user input. - Boundary markers: The instructions use double quotes (
"message") to wrap the argument, providing a basic delimiter. - Capability inventory: The skill has the capability to execute shell commands via
bashon a local script. - Sanitization: There is no explicit instruction for the agent to sanitize or escape shell-sensitive characters (like backticks or
$()) within the message before interpolation into the bash command.
Audit Metadata