The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill instructs the agent to retrieve API keys from a specific configuration file at ~/.gooseworks/credentials.json. Accessing sensitive credential files is a security-sensitive pattern, even when used for legitimate tool authentication.
[COMMAND_EXECUTION]: The setup process involves executing shell commands via python3 -c to parse local JSON files and export environment variables, which demonstrates the use of local code execution for configuration.
[EXTERNAL_DOWNLOADS]: The skill utilizes npx (specifically npx gooseworks login and npx goose-skills install) to download and execute packages from the npm registry, which introduces external code into the environment.
[PROMPT_INJECTION]: The skill's core functionality is designed to ingest large amounts of content from the web via search, extraction, and crawling. This provides a significant surface for indirect prompt injection attacks.
Ingestion points: Data enters the agent's context from arbitrary third-party URLs processed by the /search, /extract, /map, and /crawl endpoints.
Boundary markers: The instructions do not define any boundary markers or special delimiters to help the agent distinguish between its instructions and retrieved content.
Capability inventory: The skill possesses capabilities for network operations (curl) and local command execution (python3, npx).
Sanitization: There is no evidence of content sanitization or safety filtering applied to the results retrieved from the web before they are processed by the agent.

web-search-tavily