ai-seo

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

• Potentially malicious external URL detected (high risk: 0.90). The provided runner (scripts/run.py) makes runtime POST requests to the external endpoint https://ai-skills.ai/api/execute (default AISKILLS_BASE_URL) and depends on that remote service to execute the skill and return instructions/results that directly control the agent's behavior.