The Agent Skills Directory

[SAFE]: The skill implements legitimate functionality for discovering and downloading 3D assets from a reputable source.
[EXTERNAL_DOWNLOADS]: The skill fetches asset metadata and archive files from ambientcg.com via their official v3 API. These are well-known public domain resources, and the network operations are limited to this domain.
[COMMAND_EXECUTION]: No shell commands, subprocess spawning, or arbitrary code execution patterns were found. The skill relies on standard Python libraries (urllib, shutil, pathlib) for its operations.
[DATA_EXFILTRATION]: No access to sensitive files, credentials, or environment variables was detected. Network activity is strictly scoped to the asset discovery and download process.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted metadata (such as asset titles and tags) from the external API. However, it handles this data as structured JSON and returns it to the agent context without performing unsafe operations, representing a minimal risk surface typical for data-ingestion skills.

ambientcg-assets