The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses curl and python3 for standard API communication and JSON parsing. These actions are limited to the skill's stated purpose of retrieving analytics data.
[DATA_EXFILTRATION]: Network traffic is restricted to trusted Google domains (google.com, googleapis.com). No unauthorized data transmission to unknown third parties was found.
[INDIRECT_PROMPT_INJECTION]: The skill ingests external data from Google Analytics reports, representing a theoretical surface for data-driven instruction injection.
Ingestion points: Data returned from the GA4 API in SKILL.md workflows.
Boundary markers: None.
Capability inventory: Shell access for curl and network connectivity.
Sanitization: The skill performs standard JSON parsing but lacks specific sanitization for untrusted content.
[DYNAMIC_EXECUTION]: Static Python snippets are used for response formatting. The execution is localized and does not involve untrusted remote code.

google-analytics