The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches data from an external domain (northquaybooks.com) to audit sitemaps and makes requests to the Google Search Console API (googleapis.com). These downloads are used to gather data for analysis and do not involve executable code.
[COMMAND_EXECUTION]: The instructions provide shell commands using curl, grep, sed, and wc to parse XML sitemaps. It also includes a Python script (gsc_audit.py) that the agent is expected to execute locally to process JSON data from Search Console.
[INDIRECT_PROMPT_INJECTION]: The skill's primary workflow involves ingesting untrusted data from multiple sources, including support tickets, live-chat logs, sales-call notes, and Search Console API responses. This creates a surface where adversarial instructions embedded in the data could influence agent behavior. Ingestion points: local files (gsc.json) and external customer logs; Capabilities: file system access and network operations (curl); Boundary markers: not explicitly defined in the instructions; Sanitization: the provided Python script performs basic JSON parsing but does not include validation or sanitization of the content fields.

content-strategy