sag
Pass
Audited by Gen Agent Trust Hub on Aug 1, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill installs the
sagutility from a third-party Homebrew tap (steipete/tap/sag). - [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands using the
sagbinary for generating and playing audio. - [PROMPT_INJECTION]: The skill generates commands by interpolating user-controlled text directly into shell arguments (e.g.,
sag -v Clawd -o /tmp/voice-reply.mp3 "Your message here"). This represents an indirect prompt injection surface where malicious input could attempt to escape the command string. -- Ingestion points: User requests for voice responses processed inSKILL.md. -- Boundary markers: None present to distinguish user content from command arguments. -- Capability inventory: The skill leverages shell command execution of thesagbinary. -- Sanitization: The instructions do not specify any sanitization or escaping of the user-provided text before shell execution.
Audit Metadata