The Agent Skills Directory

[PROMPT_INJECTION]: The skill facilitates news monitoring by ingesting data from external platforms like Google News, Hacker News, and Reddit (see references/newsjacking.md). This creates a surface for indirect prompt injection.
Ingestion points: Untrusted data enters the agent context through shell-based retrieval of news feeds.
Boundary markers: None explicitly defined for the fetched content.
Capability inventory: Uses curl, xmllint, and jq for processing data.
Sanitization: No specific sanitization or filtering of the fetched headlines is described. This behavior is consistent with the skill's primary research purpose.
[COMMAND_EXECUTION]: The skill documentation includes examples of using common command-line utilities (curl, jq, xmllint) to fetch and parse public RSS and JSON feeds. These operations are restricted to data retrieval for media analysis and do not include high-risk patterns like privilege escalation or persistent access.

public-relations