sf-implement

Pass

Audited by Gen Agent Trust Hub on Jun 22, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill implements a rigorous implementation workflow centered on approved project artifacts and defined boundaries.
  • [COMMAND_EXECUTION]: Local Node.js scripts and Git commands are used for project management; all commands are restricted to the local project environment.
  • [DATA_EXFILTRATION]: No data exfiltration detected; network access is not requested and processing is limited to internal project files.
  • [PROMPT_INJECTION]: The instructions reinforce adherence to design documents and technical designs without any attempts to override model safety or behavior.
  • [REMOTE_CODE_EXECUTION]: No remote code is downloaded; the skill relies on local scripts and standard developer CLI tools for project setup.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 22, 2026, 02:20 AM
Security Audit — agent-trust-hub — sf-implement