The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes untrusted health data from external JSON files, which creates a potential surface for indirect prompt injection.
Ingestion points: The skill reads from various local data sources including data-example/health-goals-tracker.json, data-example/nutrition-tracker.json, and time-stamped logs in the data-example/health-goals-logs/ directory.
Boundary markers: The instructions do not define explicit delimiters or boundary markers for the data when it is interpolated into the agent's context, though they do provide processing logic.
Capability inventory: The skill utilizes Read, Grep, Glob, and Write tools. It performs calculations and generates text/HTML output but does not call dynamic execution tools like eval or exec directly on the data.
Sanitization: The documentation explicitly instructs the agent that user-derived text must be escaped and data must be safely serialized to prevent the insertion of executable scripts or malicious HTML into the final reports.
[EXTERNAL_DOWNLOADS]: The skill's report generation function references the ECharts 5.x library via a Content Delivery Network (CDN) to enable interactive visualizations.

goal-analyzer