latticework

Pass

Audited by Gen Agent Trust Hub on Aug 22, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is entirely instructional and does not include any scripts, shell commands, or executable binaries. All logic is provided as natural language instructions for the agent.
  • [SAFE]: There are no external dependencies or package managers (npm, pip) invoked by the skill.
  • [SAFE]: All external links point to the author's official infrastructure (deciqai.com, github.com/deciqai) or established academic and educational sources such as Stanford HAI and arXiv. These are used for reference purposes and do not involve remote code execution.
  • [SAFE]: No patterns of prompt injection, data exfiltration, persistence, or obfuscation were detected.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 22, 2026, 08:15 AM
Security Audit — agent-trust-hub — latticework