Skills
Navigate
skills/hatemhosny/diagrams-js/diagrams-js/json-serialization/Gen Agent Trust Hub

diagrams-js/json-serialization

Pass

Audited by Gen Agent Trust Hub on Jul 31, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill documents the fromJSON method, which dynamically loads provider modules based on the input JSON. This dynamic loading is a core functionality of the library for resolving service-specific icons and metadata.
  • [PROMPT_INJECTION]: The skill processes external JSON data, creating an indirect prompt injection surface.
  • Ingestion points: Diagram.fromJSON in SKILL.md.
  • Boundary markers: Absent in provided examples.
  • Capability inventory: Dynamic module loading and file writes (save(), writeFileSync).
  • Sanitization: Structural validation is mentioned as a prerequisite for successful parsing.
  • [DATA_EXPOSURE]: The skill demonstrates reading and writing diagram files to the local file system using fs.readFileSync and fs.writeFileSync, which is appropriate for its purpose.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 31, 2026, 01:02 AM
Security Audit — agent-trust-hub — diagrams-js/json-serialization