The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from Google Search Console, such as page paths and annotation labels, which could potentially contain malicious instructions. * Ingestion points: Untrusted data enters the agent context via get_decay_overview (page dimensions) and list_annotations (event labels) as specified in SKILL.md. * Boundary markers: The instructions do not define clear delimiters or provide 'ignore embedded instructions' warnings for the data retrieved from external tools. * Capability inventory: The skill possesses the create_annotation capability for writing data, but lacks dangerous system-level capabilities such as arbitrary shell execution or file system writes. * Sanitization: No sanitization, filtering, or validation is implemented for the text strings returned by the tools before they are used in recommendations.

traffic-forecast