networkx
Pass
Audited by Gen Agent Trust Hub on Oct 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
- [SAFE]: The skill provides standard documentation and code snippets for the NetworkX Python library (v3.7). No malicious code, unauthorized exfiltration, or obfuscated instructions were detected. All external URLs point to reputable sources such as official documentation, GitHub organizations, and arXiv.\n- [INDIRECT_PROMPT_INJECTION]: Surface Analysis:\n
- Ingestion points: The skill provides multiple methods for reading graph data from external files (references/io.md), including GraphML, GML, JSON, and CSV, as well as SQL databases.\n
- Boundary markers: Not applicable, as the data is processed programmatically for graph construction rather than being interpolated into natural language prompt templates.\n
- Capability inventory: The skill includes file system write capabilities for saving visualizations and graph data, and SQL database interaction via sqlite3 and pandas.\n
- Sanitization: The skill proactively instructs the user to use parameterized SQL queries and mentions sanitizing XML files before parsing to mitigate injection and XXE risks.\n- [DYNAMIC_EXECUTION]: The documentation (references/io.md) describes the use of the
picklemodule for graph persistence but includes a critical security warning advising against unpickling files from untrusted sources due to arbitrary code execution risks.
Audit Metadata