skills/labring/fastgpt/add-permission/Gen Agent Trust Hub

add-permission

Pass

Audited by Gen Agent Trust Hub on May 16, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill serves as a technical manual and developer guide for implementing authorization logic within the FastGPT framework. It provides bitmask-based permission control templates that align with standard security engineering practices.
  • [SAFE]: No evidence of prompt injection or attempts to bypass safety filters was found. The instructions are focused on technical implementation rather than altering agent behavior.
  • [SAFE]: Data handling patterns described in the authentication templates (e.g., parseHeaderCert, verify team ownership) demonstrate safe access control implementations without hardcoded credentials or unauthorized exfiltration paths.
  • [SAFE]: There are no external downloads, remote script executions, or suspicious package dependencies identified in the skill files.
Audit Metadata
Risk Level
SAFE
Analyzed
May 16, 2026, 12:27 AM
Security Audit — agent-trust-hub — add-permission