The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill includes a code generation script that creates component boilerplate based on user input, creating a potential surface for indirect injection if processing untrusted data.
Ingestion points: The scripts/generate_lottie_component.py script takes input via command-line arguments and interactive prompts for component names and animation sources.
Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present in the generated code templates.
Capability inventory: The script has the ability to write generated source code to the local file system using the open().write() pattern.
Sanitization: User-provided inputs are interpolated directly into code templates without validation or escaping.
[DYNAMIC_EXECUTION]: The skill includes scripts/generate_lottie_component.py, which generates executable React, Vue, or Svelte source code from predefined templates. This is categorized as a low-risk script generation pattern typical of developer scaffolding tools.
[EXTERNAL_DOWNLOADS]: Documentation and examples reference external resources for libraries and animation data from well-known services.
The skill fetches the @lottiefiles/dotlottie-web library from the JSDelivr CDN.
Animation assets are retrieved from lottie.host, the official hosting platform for LottieFiles.
[COMMAND_EXECUTION]: The provided Python scripts (generate_lottie_component.py and optimize_lottie.py) perform file system operations to read, process, and save animation and component files. These operations are limited to the scripts' intended administrative and development tasks.

lottie-animations