software-email-engineering
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [COMMAND_EXECUTION]: The diagnostic script
scripts/check_email_auth.pyutilizessubprocess.runto call thedigutility for DNS lookups. The implementation uses argument arrays without a shell, which prevents command injection vulnerabilities from user-supplied domain input. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data in the form of DNS TXT records to generate reports, creating a potential surface for indirect prompt injection. 1. Ingestion points: DNS TXT record retrieval in
scripts/check_email_auth.py. 2. Boundary markers: None implemented in the script output to distinguish record content from tool output. 3. Capability inventory: Subprocess execution ofdigand stdout reporting which the agent processes. 4. Sanitization: Standard quote stripping is performed on DNS responses, but no semantic content filtering of record values is applied. - [EXTERNAL_DOWNLOADS]: The skill recommends the installation of the
dnspythonpackage viapipfor enhanced DNS resolution capabilities. This is a standard and reputable library in the Python ecosystem.
Audit Metadata