The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on executing shell commands using the uvx tool to run the edge-tts utility. This allows the agent to run external binaries and interface with the system shell.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data by interpolating placeholders into shell commands, creating an attack surface where malicious input could influence agent behavior.
Ingestion points: User-provided content is ingested via the {msg} and {filename} placeholders in SKILL.md.
Boundary markers: Absent. The skill does not use delimiters or instructions to prevent the agent from interpreting data within {msg} as instructions.
Capability inventory: The skill has the capability to execute shell commands via uvx, write files to {tempdir}, and access the network to reach Microsoft's TTS services.
Sanitization: Absent. There is no mention of escaping, validation, or filtering of the user-provided text or filename before it is passed to the shell command.
[EXTERNAL_DOWNLOADS]: The skill uses uvx, which dynamically downloads and executes the edge-tts package from the Python Package Index (PyPI). While PyPI is a standard registry, the skill does not specify a version, which could lead to non-deterministic behavior or supply chain risks if the package is compromised.

edge-tts