The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill metadata specifies a dependency on the imsg binary and provides installation instructions for an external Homebrew tap (steipete/tap/imsg). This source is not included in the trusted vendor list, representing an unverifiable third-party dependency.
[COMMAND_EXECUTION]: The skill's primary functionality is delivered through the execution of CLI commands like imsg chats, imsg history, and imsg send. This requires the agent to invoke external processes with potentially sensitive arguments.
[DATA_EXFILTRATION]: The skill provides capabilities to read private message history and send content to external phone numbers or iMessage IDs. These features can be combined to exfiltrate sensitive local data to remote entities.
[PROMPT_INJECTION]: The skill creates an attack surface for indirect prompt injection by processing untrusted incoming messages. Ingestion points: Message data is ingested via imsg history and imsg watch commands. Boundary markers: There are no defined boundary markers or instructions to treat message content as data rather than instructions. Capability inventory: The skill possesses the imsg send capability, which allows it to transmit data externally. Sanitization: The skill does not implement sanitization or filtering of the message body before it enters the agent's context.

imsg