skills/steipete/clawdis/imsg/Gen Agent Trust Hub

imsg

Pass

Audited by Gen Agent Trust Hub on Mar 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes the imsg command-line tool to perform actions such as listing chats, retrieving message history, and sending text messages or attachments.
  • [EXTERNAL_DOWNLOADS]: The skill instructions include installing the imsg tool via a Homebrew tap (steipete/tap/imsg). This is a vendor-owned resource provided by the author of the skill.
  • [DATA_EXFILTRATION]: The skill accesses sensitive personal data stored in the macOS iMessage database, including contact details and private conversation history. Use of this functionality requires granting the terminal 'Full Disk Access' and 'Automation' permissions on macOS.
  • [PROMPT_INJECTION]: Reading incoming message content creates an attack surface for indirect prompt injection from external senders. 1. Ingestion points: iMessage history and chat data accessed via imsg history and imsg chats (SKILL.md). 2. Boundary markers: None present to distinguish untrusted message content from system instructions. 3. Capability inventory: Capability to send messages via imsg send (SKILL.md). 4. Sanitization: No content validation or escaping is applied to message data before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 15, 2026, 09:08 AM