The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external email threads, creating a surface for indirect prompt injection attacks where malicious instructions embedded in emails could influence the agent's actions.
Ingestion points: The Superhuman_Mail.get_thread tool is used in the 'Handling scheduling threads' section to read email content.
Boundary markers: Absent. The instructions do not provide delimiters or warnings to the agent to disregard instructions found within the ingested email content.
Capability inventory: The skill has powerful capabilities including Superhuman_Mail.create_or_update_event, Superhuman_Mail.create_or_update_draft, and Superhuman_Mail.send_draft which can be triggered based on thread analysis.
Sanitization: Absent. There is no mention of filtering, escaping, or validating the text retrieved from email threads before the agent uses it to decide on scheduling logic or drafting replies.

meeting-scheduler