The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes external email content, creating a vulnerability surface for indirect prompt injection where a third-party sender could attempt to influence agent behavior. \n- Ingestion points: Email threads are read using the paperclipai email thread command and the /api/companies/{companyId}/email/tasks/{taskId} API endpoint as described in SKILL.md. \n- Boundary markers: The instructions do not define specific delimiters or guidelines to distinguish email content from instructions. \n- Capability inventory: The skill allows the agent to send emails, reply to messages, and access task attachments. \n- Sanitization: There is no mention of filtering or sanitizing email bodies before processing. \n- [COMMAND_EXECUTION]: The skill instructions direct the agent to utilize the paperclipai command-line interface for email operations. \n- [DATA_EXFILTRATION]: The skill enables the agent to send data to external recipients via email, which is its primary intended function but represents a data exfiltration vector if abused.

agentmail