The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes the infsh command-line tool via Bash to interact with the inference.sh API. This is the intended functionality of the skill. The allowed-tools manifest correctly limits the agent to the infsh command, reducing the risk of arbitrary command execution.
[EXTERNAL_DOWNLOADS]: The skill references the infsh CLI as a dependency and provides installation instructions via npx. This is a standard practice for utilizing the inference.sh service ecosystem.
[DATA_EXFILTRATION]: No evidence of unauthorized data collection or exfiltration was found. Network activity is directed toward the official inference.sh domain for the purpose of speech synthesis.
[PROMPT_INJECTION]: The skill instructions do not contain any bypass attempts or overrides of the agent's safety guidelines. The text input is used exclusively for audio generation.

text-to-speech