The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the belt CLI tool to execute remote application runs on the inference.sh platform.
[EXTERNAL_DOWNLOADS]: The skill references and fetches configuration guidelines and additional skills from the inference-sh organization on GitHub, which is the official repository for the platform's tooling.
[DYNAMIC_EXECUTION]: The skill's primary function is to render user-provided React/Remotion component code (code parameter) into video files. This execution occurs within the remote inference.sh infrastructure.
[INDIRECT_PROMPT_INJECTION]: The skill defines an attack surface for indirect injection by ingesting untrusted TSX code and props.
Ingestion points: The code and props fields in the input schema for the remotion-render app.
Boundary markers: No explicit markers or instructions to ignore embedded commands are present in the skill definition.
Capability inventory: The skill uses the belt CLI tool to transmit data to a remote rendering service; it does not directly perform local file writes or arbitrary shell execution outside of the limited belt scope.
Sanitization: No specific sanitization or validation logic for the input TSX code is documented in the skill instructions.

remotion-render