create-speech-recording-desktop
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the
agent-voicecommand-line interface to perform speech synthesis tasks. - Evidence:
agent-voice speak "$RESPONSE_AS_TEXT" --waitinSKILL.md. - [EXTERNAL_DOWNLOADS]: The skill provides instructions to install and update the
agent-voicetool from the author's public repository. - Evidence:
uv tool install agent-voiceandagent-voice updateinSKILL.md. - [INDIRECT_PROMPT_INJECTION]: The skill processes text from the agent's conversation history to generate audio, creating a potential surface for indirect prompt injection if the source text contains malicious instructions.
- Ingestion points: The
$RESPONSE_AS_TEXTvariable inSKILL.mdis populated from the agent's response context. - Boundary markers: The skill instructs the agent to wrap the variable in double quotes within the shell command.
- Capability inventory: The skill has the ability to execute shell commands and write to the system temporary directory.
- Sanitization: The instructions explicitly guide the agent to convert formatting characters (e.g.,
\n,\t) into plain-text spoken forms before execution.
Audit Metadata