chrisai-graphics
Pass
Audited by Gen Agent Trust Hub on Sep 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses a provided Node.js script (
scripts/compose-brand-asset.mjs) to automate image composition tasks. The script uses internal Node.js modules (node:fs/promises,node:path) to read source images and write the resulting SVG files, serving as a functional utility for the skill's stated purpose. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data (such as source text and external assets) to generate its visual artifacts, creating a potential vulnerability surface.
- Ingestion points: Source content for graphics is ingested in
workflows/banner-and-social-graphics.mdandworkflows/infographic-design.md. - Boundary markers: The instructions do not explicitly implement delimiters or specific warnings to ignore instructions that might be embedded in the source data.
- Capability inventory: The skill possesses file read and write capabilities through the
compose-brand-asset.mjsscript. - Sanitization: The utility script implements an
escapeAttributefunction to sanitize content before it is embedded in SVG attributes, providing a mitigation against basic script or XML injection in the output files.
Audit Metadata