The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands to read local configuration and interact with APIs.
Evidence: SKILL.md contains instructions for the agent to use python3 -c to extract API keys from a local JSON file and curl to send POST requests.
[DATA_EXFILTRATION]: The skill performs network operations to an external domain that is not included in the default whitelist.
Evidence: Requests are sent to api.gooseworks.ai using curl as specified in the usage examples in SKILL.md.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes data from external, untrusted web pages.
Ingestion points: Content from external URLs is retrieved via the Olostep, Scrapegraph, and Riveter APIs defined in SKILL.md.
Boundary markers: The skill does not implement delimiters or warnings to isolate the extracted content from the agent's instructions.
Capability inventory: The skill uses curl and python3 for its core logic but does not directly execute the content it retrieves.
Sanitization: No sanitization or filtering of the ingested web data is performed before it is returned to the agent context.

extract-webpage-data