skills/huggingface/openenv/implement/Gen Agent Trust Hub

implement

Pass

Audited by Gen Agent Trust Hub on Jun 17, 2026

Risk Level: SAFE
Full Analysis
  • Automated Test Execution: The skill instructions involve finding and running failing tests using standard tools like pytest. This is a core functional component of the skill's intended use case and is performed within the user's local development environment.
  • Indirect Prompt Injection Surface: The agent is tasked with reading test files to understand requirements. While this creates a surface where the content of the tests could influence the agent's behavior, it is an inherent characteristic of code-generation tasks. In this context, it is used appropriately to guide the implementation of the requested features.
  • Minimal Privilege Usage: The skill does not request administrative privileges, perform network exfiltration, or access sensitive system configurations. It operates strictly within the project's source and test directories.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 17, 2026, 02:20 AM
Security Audit — agent-trust-hub — implement