Knowledge Graph Builder
Pass
Audited by Gen Agent Trust Hub on Feb 16, 2026
Risk Level: LOWNO_CODE
Full Analysis
- [No Code] (SAFE): Analysis of the skill reveals no executable code or scripts. The provided files consist of documentation and metadata defining an architectural framework.
- [Category 8: Indirect Prompt Injection] (LOW): The skill outlines a workflow for ingesting external data for extraction. While this introduces an attack surface in downstream implementations, the skill itself is purely instructional.
- Ingestion points: Phase 3 (Extraction) involves processing external documents to identify entities and relationships.
- Boundary markers: The instructions do not define specific delimiters or instructions to ignore embedded commands within the source data.
- Capability inventory: According to the manifest, the skill has the capacity to modify files and create artifacts.
- Sanitization: The guide does not specify validation or sanitization steps for data extracted from untrusted sources.
Audit Metadata