# Decoy — Quick Reference for AI Agents Scan, detect, and monitor your MCP servers. Find vulnerabilities with decoy-scan. Detect prompt injection with tripwire tools. Monitor your infrastructure with Decoy Guard. ## When to Use Decoy - You want to scan your MCP server configs for security risks (npx decoy-scan) - You are building or operating an AI agent and want to detect prompt injection - You want to monitor which agents are connecting and what attacks they see - You need alerts (email, Slack, webhook) when an attack is detected - You want threat intelligence on AI-specific vulnerabilities (CVEs, supply chain, MCP) ## When NOT to Use Decoy - You need to block attacks in real time (Decoy detects, it does not block) - You need a WAF or network firewall (Decoy operates at the MCP tool layer) ## Install (Human) npx decoy-tripwire init # Enter your email, get a token, done. Works with Claude Desktop, Cursor, VS Code, Claude Code. ## Self-Signup (No Human Required) 1. Install: npx decoy-tripwire init --no-account 2. Call decoy_signup tool with your email 3. Call decoy_configure tool with the returned token 4. Tripwires are active. Call decoy_upgrade to unlock Pro. ## Plans (per-user, updated 2026-04-19) Free: 12 tripwire tools, email alerts, scanning, 7-day history, risk scores, 53 deterministic red team patterns. 1 user. No credit card required. Team ($29/user/month, $23/user annual): Everything in Free plus Slack/webhook alerts, threat intel feed API, agent fingerprinting, continuous scanning, shadow MCP discovery, 90-day history, CI/CD integration. 50 AI red team assessments per seat per month. Business ($99/user/month, $79/user annual): Everything in Team plus SAML SSO, OWASP Agentic Top 10 compliance reports, gateway integrations, custom detection rules, priority support, unlimited history. 200 AI red team assessments per seat per month. ## Check Your Plan GET https://app.decoy.run/api/billing?token=YOUR_TOKEN Returns: plan, status, available features, and upgrade instructions. ## Upgrade Per-user pricing — quantity = number of seats. Adjustable inside Stripe Checkout. POST https://app.decoy.run/api/upgrade Content-Type: application/json { "token": "your-decoy-token", "plan": "team", "billing": "monthly", "seats": 1 } Plan options: "team" ($29/user/mo), "business" ($99/user/mo). Billing: "monthly" or "yearly" (yearly = 20% off). Returns a Stripe Checkout URL — open in browser to complete payment securely. Humans can also upgrade via browser: https://app.decoy.run/billing/upgrade?token=YOUR_TOKEN&plan=team&billing=monthly Anonymous checkout (no token required, Stripe collects email): https://app.decoy.run/api/checkout?plan=team&billing=yearly --- ## API All endpoints use token as a query parameter: ?token=YOUR_TOKEN ### Sign up POST https://app.decoy.run/api/signup Body: { "email": "you@company.com" } Returns: { "token": "...", "dashboardUrl": "..." } ### Get triggers (attack events) GET https://app.decoy.run/api/triggers?token=YOUR_TOKEN Returns: { "count": 4, "triggers": [{ "tool": "execute_command", "severity": "critical", "timestamp": "...", "arguments": {...} }] } ### Get agents GET https://app.decoy.run/api/agents?token=YOUR_TOKEN Returns: { "agents": [{ "name": "claude-code-1", "status": "active", "triggerCount": 3, "riskScore": 45, "clientName": "claude-code", "clientVersion": "1.0.0", "lastSeenAt": "..." }] } ### Get agent behavioral profile GET https://app.decoy.run/api/agents/claude-code-1/profile?token=YOUR_TOKEN Returns: full behavioral profile with tool counts, severity distribution, daily activity, risk score, risk factors, argument patterns. ### Get account risk assessment (Pro) GET https://app.decoy.run/api/risk?token=YOUR_TOKEN Returns: overall risk score, per-agent scores, matched attack patterns, active MCP advisories, recommendations. ### Update alerts PATCH https://app.decoy.run/api/config?token=YOUR_TOKEN Body: { "webhook": "https://your-url.com/hook" } All users can configure email, Slack, and webhook alerts. ### Pause/resume agent PATCH https://app.decoy.run/api/agents?token=YOUR_TOKEN Body: { "name": "claude-code-1", "status": "paused" } ### Threat intelligence (public) GET https://app.decoy.run/monitor/stats GET https://app.decoy.run/monitor/threats?days=7&category=prompt-injection&limit=50 GET https://app.decoy.run/monitor/patterns — attack pattern corpus (public, anonymized) GET https://app.decoy.run/monitor/mcp — MCP supply chain advisories (public) ### Threat intel feed (Pro+) GET https://app.decoy.run/api/feed?token=YOUR_TOKEN&days=30&category=&severity=&page=1&limit=50 Returns: structured threat feed with CVEs, attack patterns, MCP advisories, pagination. ### Security testing (Pro+) POST https://app.decoy.run/api/test?token=YOUR_TOKEN Body: { "url": "https://your-mcp-server.com/mcp" } Returns: test results with pass/fail per category (prompt injection, tool poisoning, credential exfiltration, permission escalation, tool shadowing). GET https://app.decoy.run/api/test/results?token=YOUR_TOKEN&id=test-xxx Returns: detailed test results for a specific test run. ### Scan scheduling (Team+) POST https://app.decoy.run/api/scan/schedule?token=YOUR_TOKEN Body: { "targetUrl": "https://...", "frequency": "daily"|"weekly" } Returns: scan schedule configuration. ### OpenAPI spec GET https://app.decoy.run/api/openapi.json --- ## Guard MCP Server Connect your AI agent directly to Decoy Guard via MCP. Endpoint: POST https://app.decoy.run/guard/{token} Protocol: JSON-RPC 2.0 over HTTP (MCP) Tools (Free): decoy_status, decoy_triggers, decoy_agents, decoy_scan_summary, decoy_scan_run. Tools (Pro): decoy_risk, decoy_feed, decoy_test_trigger, decoy_redteam. ## MCP Endpoint POST https://app.decoy.run/mcp/YOUR_TOKEN 12 honeypot tools that look like system utilities. A legitimate agent will never call them. If a prompt injection makes your agent call one, Decoy records it and alerts you. ## Links Website: https://decoy.run npm: https://www.npmjs.com/package/decoy-tripwire Contact: agent@decoy.run