ai-streaming
Pass
Audited by Gen Agent Trust Hub on Jul 27, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill offers legitimate instructional content and code samples for building AI-powered features using industry-standard tools and patterns.
- [COMMAND_EXECUTION]: Uses basic shell commands (
cat,grep,find) in Phase 1 to analyze the project's current state and dependencies. These operations are benign and restricted to project discovery. - [EXTERNAL_DOWNLOADS]: Recommends the installation of official, well-known packages from the Vercel AI SDK ecosystem (
ai,@ai-sdk/anthropic,@ai-sdk/openai) via the standard package manager. - [SAFE]: Provides security-conscious guidance for credential management by instructing users to add placeholders to a
.env.examplefile rather than hardcoding actual keys, which is a recognized best practice.
Audit Metadata