The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted external data, such as email headers, message content, and URLs, which may contain malicious instructions or payloads designed to influence agent behavior.
Ingestion points: The skill processes full original headers, email content, URLs, and attachments as described in the 'Method' section of SKILL.md.
Boundary markers: The instructions advise the agent to 'defang and detonate links in a sandbox' and to use specialized triage contexts for attachments, which provides a layer of isolation.
Capability inventory: The skill utilizes network diagnostic tools including dig and nslookup, as well as the urlscan utility for external lookups.
Sanitization: The methodology recommends the use of external sandbox services (urlscan.io) and dedicated malware triage pipelines rather than direct interaction with suspicious content.

network-email-security