The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill includes utility scripts that facilitate the processing of external JSON data, creating a potential ingestion path for indirect prompt injection attacks.
Ingestion points: The scripts/analyze_onboarding_metrics.py script ingests analytics.json, and scripts/generate_tour_config.js processes user-flows/onboarding.json as part of their standard execution flow.
Boundary markers: There are no explicit delimiters or boundary instructions provided in the skill documentation or script stubs to isolate untrusted data.
Capability inventory: The skill provides instructions for the agent to execute local Python and Node.js scripts which can perform file system operations.
Sanitization: The current script implementations are functional stubs and do not yet include input validation, schema enforcement, or data sanitization logic to handle potentially malicious payload content.

guiding-users