social-card-gen
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input from local files and external URLs, which creates a surface for indirect prompt injection where malicious instructions embedded in the source text could attempt to manipulate the agent's behavior.
- Ingestion points: Data enters via the
--text,--file, and--urlarguments ingenerate.js. - Boundary markers: The skill interpolates input directly into platform-specific templates (as seen in
composeContextingenerate.js) without implementing distinct markers to isolate the untrusted content. - Capability inventory: The skill uses
node:fsfor reading and writing files and the globalfetchAPI for retrieving external content. - Sanitization: While the script includes a
stripMarkdownAndHtmlfunction, it is designed for content formatting rather than security sanitization against prompt injection patterns. - [EXTERNAL_DOWNLOADS]: The
generate.jsscript includes functionality to fetch content from user-specified remote URLs using the Node.jsfetchAPI.
Audit Metadata