email-quality-auditor

Pass

Audited by Gen Agent Trust Hub on Aug 20, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses git rev-parse --show-toplevel to resolve the project root environment variable (AARON_SKILLS_ROOT). This is a standard and safe practice for relative path resolution in a development environment.
  • [DATA_EXPOSURE]: The skill reads and writes to specific audit memory paths (memory/audits/email/). This is limited to generated audit artifacts and requires explicit user authorization before persistence, following the least-privilege principle.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external email evidence (headers, DNS, engagement reports). It mitigates risks by requiring these inputs to be normalized and compared against a 'truth set' rather than executing instructions contained within the data. It focuses on auditing structured metrics rather than interpreting natural language instructions from the evidence.
  • [REMOTE_CODE_EXECUTION]: The instructions explicitly forbid fetching mutable remote sources (e.g., 'never fetch mutable main') and require the use of bundled immutable reference files (references/auditor-runtime.md) for standalone installations, which prevents runtime code injection or unexpected updates.
  • [SAFE]: The skill includes clear 'veto' criteria for email safety (e.g., broken authentication, unlawful lists, non-functional opt-outs) and mandates that no email sending or provider mutation occurs without separate, explicit approval.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 20, 2026, 07:28 PM
Security Audit — agent-trust-hub — email-quality-auditor