The Agent Skills Directory

REMOTE_CODE_EXECUTION (CRITICAL): The skill contains the command curl -fsSL https://cli.inference.sh | sh. This is a high-risk pattern that downloads and executes code from a remote, untrusted server without any integrity checks or verification.\n- EXTERNAL_DOWNLOADS (HIGH): The skill facilitates downloading the infsh CLI and additional components via npx from sources that are not included in the trusted provider list, increasing the risk of supply chain attacks.\n- COMMAND_EXECUTION (LOW): The skill utilizes the Bash tool to run infsh commands, which is expected for its functionality but grants the skill the ability to execute arbitrary CLI operations.\n- PROMPT_INJECTION (LOW): The skill presents an indirect prompt injection surface (Category 8) where external or untrusted data could be interpolated into the prompt argument of the TTS engine. Evidence: (1) Ingestion Point: prompt field in the JSON input to infsh app run. (2) Boundary Markers: Minimal markers (JSON formatting). (3) Capability Inventory: Bash tool usage. (4) Sanitization: No evidence of input validation or escaping for the TTS prompt.

dialogue-audio