software-baas-platforms

Pass

Audited by Gen Agent Trust Hub on Aug 12, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill serves as a reference guide for evaluating backend platforms. It includes decision matrices, security model comparisons (e.g., Row Level Security in Supabase, Firebase Security Rules), and migration strategies. All instructional content is benign and adheres to standard architectural advice.
  • [SAFE]: No obfuscation, Base64-encoded strings, or hidden Unicode characters were found in any of the provided files.
  • [SAFE]: The skill does not execute any shell commands, perform remote code downloads, or interact with the file system beyond referencing its own local files. The external links provided in data/sources.json point to official documentation for reputable technology services such as Google (Firebase), Supabase, Convex, and AWS.
  • [SAFE]: There are no signs of prompt injection, bypass attempts, or instructions to override agent safety guidelines. The use of terms like "IMPORTANT" and "CRITICAL" is restricted to highlighting operational gotchas and production risks in a standard educational context.
  • [SAFE]: No hardcoded credentials, API keys, or sensitive file paths were detected. Code snippets provided in the scenarios are illustrative and use generic placeholders.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 12, 2026, 09:09 PM
Security Audit — agent-trust-hub — software-baas-platforms