tong-jincheng-interpersonal-skill

Pass

Audited by Gen Agent Trust Hub on Apr 7, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill primarily consists of descriptive markdown content, heuristics, and mindsets. It does not contain executable scripts or commands that pose a security risk to the host system.
  • [CREDENTIALS_UNSAFE]: Integration examples in TypeScript and Python correctly utilize environment variables for managing API keys (process.env.ANTHROPIC_API_KEY and os.environ.get("ANTHROPIC_API_KEY")), following security best practices to avoid hardcoding credentials.
  • [EXTERNAL_DOWNLOADS]: The skill includes an installation instruction using npx skills add hotcoffeeshake/tong-jincheng-skill. This is a documented procedure for the platform and refers to a public repository without suspicious characteristics.
  • [PROMPT_INJECTION]: The skill defines a persona and a logic framework for the agent to adopt. It does not contain instructions aimed at bypassing AI safety guardrails, overriding system prompts, or leaking internal configurations.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 7, 2026, 02:41 PM