The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands using osascript to interface with the macOS Mail.app and ripgrep (rg) to search the local filesystem. These operations are restricted to reading mail data and are consistent with the skill's documented purpose.
[INDIRECT_PROMPT_INJECTION]: The skill processes data from external sources (incoming emails), which presents a potential surface for indirect prompt injection if an email contains malicious instructions intended to be interpreted by the agent.
Ingestion points: Email headers and bodies are ingested via scripts/emlx.py and AppleScript queries in scripts/mail-across-accounts.sh.
Boundary markers: The skill does not currently use explicit delimiters or "ignore" instructions when presenting email content to the agent.
Capability inventory: The skill is read-only and lacks network-outbound or file-write capabilities (except for user-directed attachment saving), limiting the impact of any potential injection.
Sanitization: The skill utilizes the standard Python email library for MIME decoding and header unfolding, providing robust parsing of the mail format.

apple-mail