web3-ai-tools
Pass
Audited by Gen Agent Trust Hub on Apr 24, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill contains phrases such as "Ignore previous instructions" and "System prompt extraction" within a section titled "LLM Testing for AI Chatbot Targets". These are provided as examples of payloads for security researchers to use when testing target AI applications and are not attempts to manipulate the agent's own behavior.
- [EXTERNAL_DOWNLOADS]: The documentation includes instructions for users to download and install software from various GitHub repositories and package registries (npm, PyPI). This includes tools like Shannon, LuaN1ao, and the CAI Framework, which is standard for a security toolkit distribution.
- [COMMAND_EXECUTION]: The skill provides numerous shell command examples for tool configuration, environment setup, and automated vulnerability scanning (e.g., git, npm, pip, docker, and grep patterns). These are intended to be executed by the user in a controlled testing environment.
Audit Metadata