The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a Bash script (fetch_product.sh) which uses curl for network operations and python3 for processing JSON payloads and responses.
[EXTERNAL_DOWNLOADS]: The skill makes POST requests to https://ai-factory.frevana.com to retrieve Amazon product information.
[DATA_EXFILTRATION]: The skill transmits the FREVANA_TOKEN API secret to the ai-factory.frevana.com endpoint. While this is the intended authentication mechanism, the external domain is not part of the standard trusted services list.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it fetches and processes data from an external API that could contain malicious instructions.
Ingestion points: Amazon product details are retrieved from ai-factory.frevana.com and returned to the agent.
Boundary markers: The skill does not provide explicit delimiters or instructions to the agent to treat the API response content as untrusted data.
Capability inventory: The skill has the ability to write the API response to a file path specified by the user via the --output parameter.
Sanitization: The script validates that the API response is valid JSON using json.loads but does not sanitize the text content within the fields.
[CREDENTIALS_UNSAFE]: The script allows the API token to be passed as a command-line argument (--token), which can lead to credential exposure in system process logs (e.g., via ps or /proc).

amazon-product