framework-selection

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFE
Full Analysis

The skill provides architectural guidance and does not include any executable code, shell commands, or external dependencies. It describes framework features such as filesystem access and dynamic skill loading, but these are part of the documented framework's capabilities rather than actions performed by the skill itself. The analysis found no evidence of prompt injection, data exfiltration, or obfuscation. The skill is authored by a trusted vendor (langchain-ai) and aligns with standard documentation practices for their software suite.

Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 08:16 PM
Security Audit — agent-trust-hub — framework-selection