The Agent Skills Directory

[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it retrieves and processes content from untrusted external sources like email and messengers.
Ingestion points: The skill fetches data via ~~почта and ~~мессенджер (referenced in SKILL.md and reference/data_sources.md).
Boundary markers: There are no explicit delimiters or protective instructions provided to the agent to treat data from these sources as untrusted or to ignore any embedded commands.
Capability inventory: The skill includes tools for writing data to storage (~~хранилище) and sending messages (~~мессенджер), which could be abused if an injection is successful.
Sanitization: The skill does not implement sanitization or validation of the text content retrieved from external communications before presenting it to the model for summarization.

business-pulse