arize-ai-provider-integration
Pass
Audited by Gen Agent Trust Hub on May 5, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTIONCREDENTIALS_UNSAFE
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides instructions for installing the official 'arize-ax-cli' package via standard package managers (pip, uv, pipx). This is a routine installation of a tool from a well-known service.\n- [COMMAND_EXECUTION]: The skill guides the user to persist configuration by adding export statements for non-secret environment variables (like the Arize Space ID) to shell profiles (~/.zshrc,
/.bashrc) or Windows system settings. This is a standard and neutral configuration practice.\n- [CREDENTIALS_UNSAFE]: The skill accesses local .env files to retrieve configuration and stores credentials in official CLI profiles (/.arize/config.toml). The skill includes explicit security warnings to never paste raw keys in chat or use them as raw command-line flags, strictly following credential management best practices.\n- [PROMPT_INJECTION]: The skill processes integration metadata from the Arize API, presenting a surface for indirect prompt injection.\n - Ingestion points: Data returned from 'ax ai-integrations list' and 'ax ai-integrations get'.\n
- Boundary markers: Absent; API data is used directly in the context.\n
- Capability inventory: CLI command execution, file system access for configuration, and credential rotation/management.\n
- Sanitization: No specific sanitization of API-returned strings is mentioned.
Audit Metadata