networkx-social
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from external sources including text files (network.txt), GraphML files, GEXF files, and JSON files. This creates an attack surface where maliciously crafted graph metadata or node/edge attributes could attempt to influence the agent's behavior if processed as instructions.
- Ingestion points: nx.read_edgelist, nx.read_graphml, nx.read_gexf, and json.load operations within SKILL.md.
- Boundary markers: Not explicitly defined in the provided snippets.
- Capability inventory: The skill has the ability to read and write local files (network.png, graph.graphml, etc.) and perform complex graph computations.
- Sanitization: No specific sanitization or validation logic is presented for the content of the loaded graph data.
Audit Metadata