The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands to extract configuration data from a local file. Specifically, it executes python3 -c to parse JSON and retrieve API keys and base URLs from ~/.gooseworks/credentials.json.
[DATA_EXFILTRATION]: The skill performs network operations via curl to api.gooseworks.ai. These requests include the user's API key in the authorization header and send search queries to the vendor's infrastructure for processing.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and processes untrusted data from the open web through its search and content extraction capabilities.
Ingestion points: Data enters the agent context via the Search, Answer, and Get contents capabilities, which fetch external webpage text, summaries, and search results from the Exa index.
Boundary markers: The instructions do not specify any delimiters or safety warnings to help the agent distinguish between skill instructions and retrieved web content.
Capability inventory: The skill has the ability to execute shell commands (curl, python3) and perform network operations.
Sanitization: There is no evidence of sanitization or filtering of the retrieved web content before it is processed by the agent.

web-search-exa