aiden-test-feature

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt explicitly tells the agent to ask for test credentials and to use them in browser interactions (e.g., filling forms) and also directs constructing curl commands using presigned upload URLs returned by a tool, both of which require emitting secret values verbatim in generated commands—creating an exfiltration risk.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The skill mandates installing and running the external CLI "agent-browser" at runtime via "npm install -g agent-browser && agent-browser install", which fetches and executes remote code from the agent-browser package (see https://github.com/vercel-labs/agent-browser) and is required for the skill to run.