spoken-response

Pass

Audited by Gen Agent Trust Hub on Jul 28, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes shell commands using variables $TEXT, $LABEL, and $path within SKILL.md and references/recording-delivery.md. Specifically, the command printf '%s' "$TEXT" | agent-voice speak is vulnerable to shell command substitution if $TEXT contains shell metacharacters like backticks or $().
  • [EXTERNAL_DOWNLOADS]: The skill uses uv tool install agent-voice to download and install its core functionality. The package is associated with the author's official GitHub repository.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection (Category 8).
  • Ingestion points: Assistant responses are processed as input for the speak command in SKILL.md.
  • Boundary markers: The instructions do not define boundary markers to isolate processed text.
  • Capability inventory: The skill has the ability to execute shell commands and write to files.
  • Sanitization: No sanitization or escaping of the assistant's response is described before shell interpolation.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 28, 2026, 02:14 PM
Security Audit — agent-trust-hub — spoken-response