email-sequence
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill acknowledges that it processes untrusted inputs, including scraped web content, reviews, and emails. It mitigates this risk by providing clear instructions in
SKILL.mdandreferences/skill-contract.mdto treat such data exclusively as evidence and to ignore any instructions found within those sources. - [CREDENTIALS_UNSAFE]: The skill documentation in
references/connectors.mdexplicitly advises against storing API keys within skill files or source control, promoting the use of environment-based secret management. - [EXTERNAL_DOWNLOADS]: The
references/connectors.mdfile references several well-known third-party marketing, SEO, and integration platforms (such as Google Ads, HubSpot, and Zapier). These are presented as optional external data sources and represent standard industry integrations.
Audit Metadata