network-bgp-diagnostics
Pass
Audited by Gen Agent Trust Hub on Oct 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from network device outputs, creating a potential surface for indirect prompt injection.
- Ingestion points: The
parse_bgp_summaryfunction inSKILL.mdingests raw string data from network devices. - Boundary markers: The instructions do not explicitly specify the use of delimiters when passing device output to the agent's context.
- Capability inventory: The skill is restricted to read-only diagnostics (e.g.,
showcommands,ping,traceroute) and local Python string parsing. - Sanitization: The Python code utilizes specific regular expressions (
BGP_SUMMARY_RE) to validate and extract BGP fields, providing basic structural validation of the input.
Audit Metadata