venice-text-routing
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFE
Full Analysis
- [REMOTE_CODE_EXECUTION]: The helper script
scripts/refresh_routing.pyfetches model metadata from the official Venice API (api.venice.ai). This is a legitimate vendor resource used for discovering model capabilities and pricing.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes model information from the Venice API to perform routing logic.\n - Ingestion points: Model metadata is read from
snapshots/text-routing.json(populated byrefresh_routing.py).\n - Boundary markers: Model traits and identifiers are used directly in routing logic without explicit delimiters.\n
- Capability inventory: The skill's primary function is selecting model IDs for chat completions via
curl.\n - Sanitization: Metadata from the official API is processed using standard JSON parsing.\n- [COMMAND_EXECUTION]: The documentation includes standard usage examples for calling the Venice API via
curl.
Audit Metadata