The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute various reduck CLI commands to perform search, discovery, and web automation tasks. This also includes dynamic context injection in SKILL.md, where the ! syntax is used to execute help commands like reduck --help and reduck run --help during the skill loading process to inline live documentation.
[EXTERNAL_DOWNLOADS]: The instructions require the user to install the @reduck-ai/cli package via npm. This is a legitimate vendor resource provided by the skill's author to enable the functionality described.
[PROMPT_INJECTION]: The skill is designed to process content from arbitrary web pages (scraping and searching), which introduces a surface for indirect prompt injection where instructions hidden on a website could attempt to manipulate the agent's behavior.
Ingestion points: Untrusted data enters the agent context through the output of commands like reduck search <query> and reduck run --method <host>/<slug> (found in SKILL.md).
Boundary markers: The instructions suggest using structured formats like JSON or YAML for outputs, which provides some structural delimitation, but does not explicitly require the agent to ignore instructions found within the data.
Capability inventory: The agent has the capability to execute shell commands and read files (detailed in SKILL.md).
Sanitization: There is no mention of sanitization or filtering for the external web content being processed.

reduck