The Agent Skills Directory

[PROMPT_INJECTION]: The skill identifies as an analysis engine that processes untrusted external data (Amazon product titles, descriptions, and customer reviews), which creates an indirect prompt injection surface.
Ingestion points: External data is ingested via zoodata.py from the ZooData API and presented to the agent for categorization and sentiment analysis.
Boundary markers: The prompt rendering logic in scripts/zoodata.py (specifically render_review_map_prompt) uses triple-quote (""") delimiters to encapsulate review text before it is processed by the AI.
Capability inventory: The skill is capable of performing network requests to the vendor's API and executing its own Python scripts. It does not have arbitrary shell execution or file system write capabilities.
Sanitization: The skill truncates review text to 500 characters to limit the context window, though no explicit character escaping for the triple-quote delimiter is performed.
[COMMAND_EXECUTION]: The skill relies on the execution of a local Python script, scripts/zoodata.py, to communicate with the ZooData API. This script handles all data fetching and provides prompt templates for the agent.
[EXTERNAL_DOWNLOADS]: The skill performs legitimate network operations to api.zoodata.ai (the vendor's domain) to retrieve Amazon market statistics, brand details, and product information. These requests are authenticated using a user-provided API key.

amazon-analysis