Local security audit for AI API relays and LLM proxies: detects prompt injection, model substitution, tool-call rewriting, SSE anomalies, error leakage, and Web3
Local security audit for AI API relays and LLM proxies: detects prompt injection, model substitution
