The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill operates by ingesting external data points such as account metrics, support sentiment, and financial status provided by a user to generate health scores and action plans.
Ingestion points: The assessment template in SKILL.md accepts arbitrary strings for account names, feature adoption descriptions, and sentiment analysis.
Boundary markers: The prompt template lacks specific delimiters or instructions to ignore potential commands embedded within the provided data fields.
Capability inventory: The skill does not currently invoke dangerous capabilities like subprocess execution, file system writes, or network operations.
Sanitization: There are no mechanisms described for sanitizing or validating the input data before processing.

account-health