routeros-command-tree

Pass

Audited by Gen Agent Trust Hub on Sep 16, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill documentation describes a pattern for ingesting structural metadata and command descriptions from external RouterOS devices via the /console/inspect API endpoint.
  • Ingestion points: Metadata is retrieved from the /rest/console/inspect REST API via network requests as demonstrated in the TypeScript examples within SKILL.md.
  • Boundary markers: The provided code examples do not implement specific delimiters or instruction-bypass warnings for the ingested syntax descriptions.
  • Capability inventory: The documented patterns utilize network operations (fetch) to interact with device endpoints.
  • Sanitization: The implementation examples show parsing of structured JSON metadata; explicit sanitization of description text fields is not demonstrated.
  • [EXTERNAL_DOWNLOADS]: The skill references the tikoci/restraml repository on GitHub. This is a vendor-owned resource belonging to the skill author, used to provide context for command mapping and schema generation logic.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 16, 2026, 01:15 PM
Security Audit — agent-trust-hub — routeros-command-tree