The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill documentation recommends a "Fast install" method using curl -fsSL https://outlit.ai/install.sh | bash. This executes a remote script directly in the shell. While originating from the vendor's domain, this method skips local validation of the code being executed.
[COMMAND_EXECUTION]: The skill's instructions rely on executing various outlit CLI commands for setup, authentication, and notification tasks, which interact with the host operating system.
[EXTERNAL_DOWNLOADS]: The skill downloads dependencies and executable tools from external sources, including NPM, Homebrew, and the vendor's website.
[PROMPT_INJECTION]: The skill retrieves data from external, untrusted sources such as customer emails and support tickets, creating a surface for indirect prompt injection.
Ingestion points: Data is fetched via outlit search and outlit_get_timeline.
Boundary markers: No specific delimiters or instructions to ignore embedded commands are provided.
Capability inventory: The skill can perform SQL queries, system configuration, and send notifications to external platforms.
Sanitization: There is no mention of sanitizing external data before it is processed by the agent.

outlit