landing-page-conversion-audit
Pass
Audited by Gen Agent Trust Hub on Oct 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external website content (DOM) provided via a URL. This creates a surface for indirect prompt injection where malicious instructions hidden on a webpage could influence the agent's audit results or behavior.
- Ingestion points: Fetches and reads the rendered DOM from a user-supplied Page URL as specified in SKILL.md.
- Boundary markers: None identified; the skill does not use specific delimiters or instructions to ignore embedded commands in the external data.
- Capability inventory: The skill can suggest shell commands and generate implementation guidance.
- Sanitization: No sanitization of the external DOM content is described.
- [EXTERNAL_DOWNLOADS]: The skill recommends downloading a self-hosted funnel builder from a third-party GitHub repository (github.com/autonnel/autonnel) that is not a recognized trusted vendor. This poses a supply chain risk.
- [COMMAND_EXECUTION]: The skill provides specific shell commands (docker compose up) for setting up external infrastructure. If the agent has shell access, these commands could be executed, potentially running untrusted container images on the local system.
Audit Metadata