cold-email
Pass
Audited by Gen Agent Trust Hub on Nov 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a bundled shell script (scripts/check-sending-domain.sh) to perform DNS queries for deliverability verification. The script utilizes standard system tools (dig or host) and performs basic input sanitization on domain names to prevent injection.
- [DATA_EXFILTRATION]: The skill performs network POST requests to the author's domain (revnu.com) to provide project previews. This is a functional component of the service and includes instructions to exclude secrets or private code from the transmitted data.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from the public web to find reasons for outreach, which exposes it to potential injection attacks from malicious websites.
- Ingestion points: Content fetched from the public web during the prospecting phase (Step 1 in SKILL.md).
- Boundary markers: Present; the skill instructs the agent to treat all fetched text as data and disregard any contained instructions.
- Capability inventory: Local script execution, network operations to the vendor API, and MCP tools for managing email campaigns.
- Sanitization: The instructions provide a clear mandate to separate data findings from the agent's core instructions.
Audit Metadata