The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute several local scripts within the ${CLAUDE_PLUGIN_ROOT} directory using Node.js and Python 3. These scripts include tts.mjs for audio generation, trim-vo.py for audio splicing, and swap-line.py for targeted audio replacement. While these appear to be internal tools, they grant the agent the capability to run code on the local filesystem.
[INDIRECT_PROMPT_INJECTION]: The swap-line.py command processes arbitrary text provided by the user via the --find and --replace flags. This creates a surface for indirect prompt injection if the inputs are not properly sanitized before being processed by the underlying Python script.
Ingestion points: The --find and --replace arguments in the swap-line.py shell command.
Boundary markers: None are specified in the instructions to delimit user input from command logic.
Capability inventory: The skill can execute scripts (node, python3), modify audio files, and write metadata to the filesystem.
Sanitization: The instructions do not specify any validation or sanitization routines for the user-supplied text strings.

hyperframe-pro-voice