The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill documentation provides an installation command (curl -fsSL https://cli.tavily.com/install.sh | bash) that downloads and executes a shell script directly from the vendor's official domain. While this is the intended setup method for the Tavily CLI, executing remote scripts via piped shell commands is a security best practice violation.
[COMMAND_EXECUTION]: The skill uses the tvly command-line tool via a Bash interface to perform its operations. It supports writing extracted data to the local file system using the -o or --output flags.
[EXTERNAL_DOWNLOADS]: The skill is designed to fetch and process content from arbitrary third-party URLs provided by the user.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests untrusted data from the web and presents it to the agent.
Ingestion points: Web content retrieved from URLs passed to the tvly extract command.
Boundary markers: No specific delimiters or instructions are used to separate external content from the agent's system instructions.
Capability inventory: The skill has access to the Bash tool, which can execute the tvly CLI and write files to the local disk.
Sanitization: Content is processed into clean markdown, but there is no evidence of filtering or sanitizing the text to prevent embedded instructions from influencing the agent's behavior.

tavily-extract