The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill metadata and instructions direct the installation of a Go binary from a remote repository (github.com/Hyaxia/blogwatcher). This constitutes a download of external executable code from a third-party source.\n- [COMMAND_EXECUTION]: The skill functions by executing the blogwatcher CLI tool on the system. Commands such as scan, add, and read involve local process execution to manage feed data.\n- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection (Category 8) due to its core function of processing external web data.\n
Ingestion points: External data enters the agent context through blogwatcher scan and blogwatcher articles, which fetch and display content from RSS/Atom feeds.\n
Boundary markers: Absent; the skill does not utilize delimiters or specific instructions to prevent the agent from following commands embedded within the blog articles.\n
Capability inventory: The skill facilitates command execution and network requests via the CLI tool, which could be abused if malicious instructions are ingested.\n
Sanitization: Absent; there is no evidence of sanitization or filtering applied to the fetched feed content before it is processed by the agent.

blogwatcher