self-hosted-funnel-launch
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill facilitates the deployment of the Autonnel platform by providing links to its official GitHub repository and container images on GitHub Container Registry (GHCR). These are established vendor resources for the 'autonnel' project.
- [INDIRECT_PROMPT_INJECTION]: The skill documents an interface (MCP) that allows an agent to create and update funnel content. It correctly identifies a potential security surface by noting that the tool does not structurally validate certain content fields before saving, which could be exploited if an agent processes untrusted external data.
- Ingestion points: The
create_pageandupdate_pageMCP tools described in the 'operate it from an agent over MCP' section. - Boundary markers: The skill provides specific rules and schemas for agents to follow, though it notes the underlying system's lack of enforcement.
- Capability inventory: Capabilities include writing web content (HTML/JSON), configuring payment funnels, and managing media assets.
- Sanitization: The documentation explicitly cautions that "Nothing checks root / content / zones... before saving," highlighting the need for careful data handling.
Audit Metadata