ecommerce-tiktok-video-rank
Pass
Audited by Gen Agent Trust Hub on Sep 11, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill relies on a local Python script (
scripts/echotik_list_video_rank.py) to interact with the external API. The agent is instructed to execute this script using the command line with JSON-encoded parameters. This is the intended primary functionality of the skill. - [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external TikTok video rankings, which includes untrusted content like video descriptions and nicknames. This creates an attack surface where malicious content from the API could attempt to influence the agent's behavior during the summarization phase. The skill mitigates this by providing structured instructions for summarizing only returned facts and preserving missing values.
- [DATA_EXPOSURE]: The skill manages authentication via the
NEXSCOPE_API_KEYenvironment variable. It explicitly instructs the user not to paste credentials into the chat or JSON parameters, which aligns with security best practices. Network communication is directed to vendor-controlled domains (api.nexscope.aiandskill-api.nexscope.com). - [PERSISTENCE]: The Python script implements a caching and logging mechanism that writes API responses to a
nexscopedirectory within the workspace or temporary folders. This behavior is documented as part of the skill's data preservation workflow and does not involve malicious persistence mechanisms like modifications to shell profiles or cron jobs.
Audit Metadata