The Agent Skills Directory

[DATA_EXFILTRATION]: The skill accesses a sensitive credential file to retrieve an API key and base URL.
Evidence: Reads ~/.gooseworks/credentials.json to extract api_key and api_base.
[COMMAND_EXECUTION]: The skill executes shell commands and inline Python scripts to process authentication data and perform network requests.
Evidence: Uses python3 -c for parsing JSON and curl for communicating with the Gooseworks API.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection by ingesting data from external websites.
Ingestion points: Content retrieved from external websites via the scrapegraph API (e.g., https://stripe.com/pricing).
Boundary markers: None; the skill does not use delimiters or instructions to prevent the agent from obeying commands embedded in scraped content.
Capability inventory: The agent can execute shell commands (curl) and perform local Python operations.
Sanitization: No sanitization or validation of external content is performed before the data enters the agent's context.

company-intel