{"service":"Agent Safe","version":"3.0.0","description":"11-tool message security suite for AI agents (8 subscription + 3 free). Protects against phishing, BEC, malware, social engineering, manipulation, AI-generated images, deepfakes, and prompt injection across any message (emails, chats, DMs, SMS), URLs, replies, threads, attachments, sender identities, and media. Start with the free assess_message tool to get personalized recommendations, query check_prompt_injection_db for known injection patterns, and use submit_feedback to report results.","capabilities":["message_triage","email_safety_check","url_safety_check","response_safety_check","thread_analysis","attachment_safety_check","sender_reputation_check","message_safety_check","media_authenticity_check","prompt_injection_database"],"domainFocus":"message_security","pricing":{"model":"subscription_or_credits","trial":"25 free checks for every new account","plan":"$9/month personal plan (up to 3,000 checks/month fair use)","creditPacks":"500 checks/$5 or 1,500 checks/$12 — never expire","note":"Most tools use 1 unit per call; image analysis 4 units; video 10 units.","billing":"https://agentsafe.locationledger.com/billing"},"authentication":{"type":"oauth2.1","description":"Add https://agentsafe.locationledger.com/mcp as a remote MCP server — your client (Claude, ChatGPT, Gemini, Cursor) handles sign-in automatically via OAuth with dynamic client registration and PKCE.","authorizationServer":"https://agentsafe.locationledger.com/.well-known/oauth-authorization-server"},"endpoints":{"mcp":"https://agentsafe.locationledger.com/mcp","oauthMetadata":"https://agentsafe.locationledger.com/.well-known/oauth-authorization-server","billing":"https://agentsafe.locationledger.com/billing"},"documentation":"/docs","termsOfService":"/terms","termsNotice":"By making any API call or MCP request to this service, you accept the Terms of Service. This service provides best-effort advisory analysis only. We are not liable for undetected threats or actions taken by agents based on our analysis.","trustSignals":{"uptime":"99.9%","avgResponseMs":450,"agentsServed":41,"threatsBlocked":138}}