drawio-aws-diagrams

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.70). The skill explicitly instructs the agent to run a WebFetch against the public GitHub file https://github.com/jgraph/drawio/blob/dev/src/main/webapp/js/diagramly/sidebar/Sidebar-AWS4.js (see SKILL.md step 3 and STENCILS.md), meaning it ingests untrusted third-party content that directly influences stencil/name decisions and downstream diagram-generation actions.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 1.00). The skill explicitly instructs a runtime WebFetch of https://github.com/jgraph/drawio/blob/dev/src/main/webapp/js/diagramly/sidebar/Sidebar-AWS4.js to verify/add stencil names, so it fetches external content that directly controls which prompts/instructions the agent follows for stencil selection and diagram authoring.