The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted external data by fetching email subjects and body content. While the skill itself is a passive reader, the ingested content poses a risk of indirect prompt injection if a downstream agent processes or summarizes the emails without sufficient boundary markers or sanitization.
Ingestion points: The scripts/imap.js file fetches email content from an external IMAP server via connection.search and simpleParser.
Boundary markers: Absent. The script outputs raw JSON data containing email snippets and bodies.
Capability inventory: The skill is capable of reading, searching, and marking emails as read/unread.
Sanitization: None. The script does not filter or escape content from the email body or subject.
[SAFE]: The skill manages sensitive IMAP credentials using a .env file, which is a standard and recommended practice for local CLI tools to avoid hardcoding secrets. The setup.sh script correctly uses read -s to prevent the password from being echoed to the terminal during entry.
[SAFE]: The dependencies listed in package.json (imap-simple, mailparser, dotenv) are well-known, widely used, and appropriate for the skill's stated purpose.

imap-email