retrofit

Pass

Audited by Gen Agent Trust Hub on Oct 6, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides standard documentation and code templates for Android networking using Retrofit and OkHttp.
  • [CREDENTIALS_UNSAFE]: The skill correctly demonstrates using BuildConfig.BASE_URL and TokenProvider abstractions rather than hardcoding sensitive information. It includes a specific warning against hardcoding base URLs.
  • [REMOTE_CODE_EXECUTION]: No remote code execution patterns were detected. The skill uses standard libraries (Retrofit, OkHttp) at specific, well-known versions.
  • [COMMAND_EXECUTION]: The skill does not use any shell command execution or subprocess spawning.
  • [EXTERNAL_DOWNLOADS]: The skill references standard dependencies in a libs.versions.toml format targeting well-known Maven repositories (Squareup, JetBrains), which are trusted sources.
  • [PROMPT_INJECTION]: No prompt injection or behavior override patterns were detected in the instructions.
  • [DATA_EXFILTRATION]: The skill demonstrates standard API interactions. While it handles authentication tokens, it does so through standard Android patterns (Interceptors, Authenticators) without exfiltrating data to untrusted domains.
  • [INDIRECT_PROMPT_INJECTION]: While the skill handles external API data, it recommends using kotlinx.serialization with ignoreUnknownKeys = true and wrapping results in a sealed NetworkResult class, which are standard safety practices for handling untrusted external data.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 6, 2026, 04:16 PM