cognitive-science-landscape

Pass

Audited by Gen Agent Trust Hub on Aug 18, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns, obfuscation, or persistence mechanisms were detected in the skill files.
  • [EXTERNAL_DOWNLOADS]: The skill references external URLs including deciqai.com, github.com/deciqai, archive.org, and doi.org. These are legitimate vendor resources or reputable academic repositories used for documentation and sourcing citations.
  • [COMMAND_EXECUTION]: The skill does not contain any shell commands, script executions, or system-level operations.
  • [PROMPT_INJECTION]: The skill contains instructional logic for a coaching mode using [WAIT] markers. These are benign interaction design patterns intended to guide the AI's conversational flow and do not represent a security risk.
  • [DATA_EXPOSURE]: There are no hardcoded credentials, sensitive file paths, or data exfiltration patterns identified in the skill.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 18, 2026, 05:57 AM
Security Audit — agent-trust-hub — cognitive-science-landscape