neo4j-gds-skill
Pass
Audited by Gen Agent Trust Hub on Apr 28, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides documentation and code snippets for official Neo4j Graph Data Science (GDS) procedures and the
graphdatasciencePython client. All identified URLs and packages originate from the vendor or trusted sources.\n- [PROMPT_INJECTION]: Analysis of indirect prompt injection surface (vulnerability surface identification):\n - Ingestion points: Graph properties and structures are read from the Neo4j database into in-memory graphs via
gds.graph.projectorgds.graph.cypher.project(SKILL.md).\n - Boundary markers: None; the skill assumes graph data follows expected structural schemas.\n
- Capability inventory: The skill facilitates executing various GDS algorithms (centrality, community detection, path finding) and writing results back to the database using
write-cyphercommands.\n - Sanitization: No content-based validation of graph property values is performed prior to algorithm execution.
Audit Metadata