dial-your-context

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly instructs the agent to call the Sanity Context MCP endpoint (e.g. https://api.sanity.io/vX/context/mcp/{project}/{dataset}/{slug}?instructions=&groqFilter=) or write/read the context doc at runtime, and the returned "instructions" blob from that URL is injected into the agent's prompt, so remote content directly controls agent instructions.

MEDIUM W021: Hidden or invisible Unicode characters detected (potential obfuscation or prompt injection).

• Hidden Unicode characters detected (1 type(s) found)