li-er
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [No Code] (SAFE): The skill is entirely composed of Markdown text and instructional prompts. No executable code files (.py, .js, .sh), binary executables, or package manifests (package.json, requirements.txt) are present, eliminating the risk of command execution or system compromise.
- [Prompt Injection] (SAFE): The core of the skill is a system prompt defining a deceptive and evasive persona. While the instructions mandate the AI to adopt a satirical, 'lying' communication style within the roleplay, this is context-specific and does not attempt to bypass the underlying AI safety filters, leak system configurations, or override the agent's primary directives.
- [Data Exposure & Exfiltration] (SAFE): No hardcoded credentials, API keys, or sensitive file paths (such as .ssh or .aws configs) were detected in the skill files. The skill does not possess the capability to perform network requests or exfiltrate data.
- [External Links] (SAFE): The skill includes a promotional link to an external website (www.k-dense.ai). This is a static reference to a web-based platform and does not involve automated downloads, piped script execution, or remote code loading.
Audit Metadata