agent-ready-cloudflare

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). In SKILL.md Step 2 scans a user-supplied target URL (via POST to https://isitagentready.com/api/scan), and then Step 4 embeds the API response field checks...message (“Issue”) into “How to Implement” prompts, meaning text originating from the scanned website becomes LLM-visible through the API response.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill calls the external scan API at https://isitagentready.com/api/scan (and optionally the MCP endpoint https://isitagentready.com/mcp) at runtime and uses the API response fields (messages/prompts/nextLevel.prompt) to construct and control the fix prompts, so the external content directly controls agent output.

MEDIUM W009: Direct money access capability detected (payment gateways, crypto, banking).

• Direct money access detected (high risk: 1.00). The skill explicitly includes commerce/payment protocols and implementation guidance (x402, MPP, UCP, ACP, AP2). Sections mention concrete payment middleware, facilitator URL and wallet address, and MPP methods (stripe, lightning, card) with language about agents fulfilling payments automatically. These are specific payment integrations (including wallet/crypto and payment gateway methods), not generic tooling, so the skill exposes/directly enables agent financial execution capabilities.