The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill facilitates data retrieval from api.crawlora.net. These operations are directed to the vendor's official API endpoints and are consistent with the skill's primary purpose of researching Yahoo content.
[COMMAND_EXECUTION]: The helper script scripts/crawlora.sh executes curl to perform network requests. The script follows security best practices by validating the API key format, whitelisting specific allowed path patterns, and using a temporary configuration file to securely pass headers, preventing sensitive information from appearing in process lists.
[INDIRECT_PROMPT_INJECTION]: The skill ingests article content and reader comments from external sources, which represents a potential injection surface.
Ingestion points: Untrusted content is ingested via the /yahoo-news/comments, /yahoo-news/article, and other vertical-specific article endpoints in scripts/crawlora.sh.
Boundary markers: The instructions do not specify the use of delimiters or specific warnings to the agent regarding the handling of the fetched content.
Capability inventory: The skill is limited to performing network GET requests and outputting the results to the agent's context. It does not have permissions to write to the permanent filesystem or execute arbitrary code.
Sanitization: The script returns the API's JSON response directly; no sanitization or filtering of the article body or comment text is performed before it is provided to the agent.

yahoo-network-research