list-hygiene-monitor
Pass
Audited by Gen Agent Trust Hub on Sep 2, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill invokes local Python scripts (
resend.pyandledger.py) located within the platform's plugin directory to fetch contacts and record KPIs. These are functional calls to internal tools required for the skill's primary purpose. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data sources, specifically ESP engagement exports and bounce/complaint reports, which creates a potential attack surface.
- Ingestion points: Subscriber engagement exports, bounce reports, and suppression dumps (SKILL.md).
- Boundary markers: Present. The instructions explicitly mandate treating all exported files as untrusted and state that text within an export must be treated as data, never as a command (SKILL.md).
- Capability inventory: The skill uses shell commands to execute local scripts and has the ability to write reports and summaries to the
memory/directory (SKILL.md). - Sanitization: While the markdown instructions do not define programmatic escaping, they provide clear behavioral guidelines for the agent to differentiate between data and instructions.
Audit Metadata