The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill documentation specifies an installation command curl -fsSL https://cli.inference.sh | sh which pipes a remote script directly into the shell. This is a high-risk execution pattern as it relies on the integrity of the remote server and the transport layer.
[EXTERNAL_DOWNLOADS]: The skill's installation process downloads binary executables from dist.inference.sh to provide the core functionality. These external files are required for the skill's primary operation.
[COMMAND_EXECUTION]: The skill is configured to use the Bash tool to run infsh commands, allowing the agent to execute subcommands via the external CLI tool.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection via the code parameter. 1. Ingestion points: The code parameter in the infsh app run command accepts arbitrary TSX code that could originate from untrusted user input. 2. Boundary markers: The input is encapsulated within a JSON string, providing a structural boundary but no behavioral constraints against instructions embedded in the code. 3. Capability inventory: The skill uses Bash to execute the infsh CLI which performs rendering. 4. Sanitization: No explicit validation or sanitization is performed on the input code within the skill definition.

remotion-render