skills/benknightdark/neo-skills/swift/Gen Agent Trust Hub

swift

Pass

Audited by Gen Agent Trust Hub on Apr 1, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns, obfuscation, or unauthorized data access detected across any analyzed files.
  • [PROMPT_INJECTION]: Instructions are purely technical and instructional, with no attempts to override agent behavior or bypass safety guardrails.
  • [DATA_EXFILTRATION]: No network access or sensitive file operations detected; external links point to official documentation.
  • [COMMAND_EXECUTION]: No shell commands or script execution logic is present in the skill content.
  • [SAFE]: Analysis of indirect injection surfaces confirms that while the skill reads external project files (Ingestion points: *.swift, Package.swift in SKILL.md), it possesses no dangerous capabilities such as shell execution or network access (Capability inventory: None), and thus presents no risk. Boundary markers and explicit sanitization are not required for this purely instructional and non-executable skill.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 1, 2026, 04:46 AM
Security Audit — agent-trust-hub — swift