The Agent Skills Directory

EXTERNAL_DOWNLOADS (MEDIUM): The skill frequently uses npx awal@latest to perform wallet and payment operations. This command downloads and executes the latest version of the 'awal' package from the npm registry at runtime. Since 'awal' is not from a trusted source, this poses a risk of executing malicious code if the package or its registry entry is compromised.
REMOTE_CODE_EXECUTION (MEDIUM): The use of npx to run unverified packages like 'awal' constitutes remote code execution. Additionally, the skill installs x402-express and @coinbase/x402, which are not on the list of trusted repositories or organizations, increasing the attack surface.
COMMAND_EXECUTION (LOW): The skill executes several shell commands (npm install, node index.js, mkdir, curl). While these are necessary for the primary purpose of the skill (deploying a server), they involve running user-generated code and managing file systems.
DATA_EXFILTRATION (SAFE): The skill mentions the use of sensitive environment variables like CDP_API_KEY_ID and CDP_API_KEY_SECRET. However, it does not attempt to access or exfiltrate these secrets; it only provides instructions for their legitimate use within an Express server setup.
DYNAMIC_EXECUTION (LOW): The skill generates a local script (index.js) and executes it using Node.js. This is the intended primary behavior of the skill and does not appear to incorporate untrusted external input into the execution logic in a malicious way.

monetize-service