multivendor-fleet-ops
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
run_fleettool to execute raw CLI commands on multiple network devices. This enables high-scale remote interaction with network infrastructure. - [INDIRECT_PROMPT_INJECTION]: The skill processes data fetched from external network devices, creating a surface for instructions embedded in device output to influence the agent.
- Ingestion points: CLI output and operational facts returned by remote devices through the
run_fleetandget_factstools (documented in SKILL.md). - Boundary markers: The instructions do not define specific delimiters or warnings to isolate device-generated text from agent instructions.
- Capability inventory: The skill provides fleet-wide command execution and data retrieval capabilities across multiple vendors.
- Sanitization: The skill documentation focuses on transport security (SSH) and access control (read-only) but does not address the sanitization of data returned to the LLM context.
Audit Metadata