live-transcribe-stop
Pass
Audited by Gen Agent Trust Hub on Jun 13, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a local bash script located at
~/.claude/scripts/stop-transcribe.shto manage the transcription process lifecycle. It also performs file system operations in the/tmp/directory to identify the most recent transcript file. - [PROMPT_INJECTION]: The skill introduces a surface for indirect prompt injection by instructing the agent to read output from a transcription process.
- Ingestion points: The agent is directed to read the content of files matching
/tmp/transcribe-*.txt. - Boundary markers: There are no specified boundary markers or instructions to ignore potential commands within the transcript.
- Capability inventory: The agent possesses the capability to execute shell commands via bash.
- Sanitization: The skill lacks mechanisms to sanitize or validate transcribed text before the agent processes it.
Audit Metadata