The Agent Skills Directory

[COMMAND_EXECUTION]: The skill is built around executing shell commands via the email CLI tool to perform mailbox operations.
[CREDENTIALS_UNSAFE]: The configuration workflow (email config --password "app-password") accepts passwords as command-line arguments, which can be logged in shell history. Furthermore, the skill's documentation explicitly notes that the CLI tool stores these passwords in plain text on the local filesystem.
[EXTERNAL_DOWNLOADS]: The skill instructions include downloading and installing the @lyhue1991/email-cli package from the NPM registry.
[PROMPT_INJECTION]: The skill processes untrusted content from external emails, which presents an indirect prompt injection surface.
Ingestion points: Incoming emails (headers and bodies) are ingested through the email receive command (found in SKILL.md).
Boundary markers: There are no mentioned delimiters or safety instructions to prevent the agent from following commands embedded in email content.
Capability inventory: The skill has the ability to send data over the network (email send) and write files to the local disk (email receive --attachments).
Sanitization: No content sanitization or validation steps are described for the received email data.

email