Pricing

Know what breaks.
Know what to fix.

Keep your bots and MCP tools under watch. Go deeper with agentic security testing, verified attack paths, and a clear plan for your engineering team.

01 / Continuous monitoring

Your bots change. Your checks should keep up.

Recurring tests for leaks, wrong answers, unsafe actions, and changing permissions. Every plan includes evidence, alerts, and guidance on what to fix.

Save 20% with annual billing
For everyday monitoring

Watch

$299/mo

Billed monthly

Know when your chatbots leak information, give the wrong answers, or stop working. Daily checks keep the basics under watch.

Start Watch
  • Up to 100 chatbot or AI application endpoints
  • Daily availability, latency, prompt leak, and safety checks
  • Weekly jailbreak and prompt injection probe pack
  • 3 canary phrases and 15 business facts checked for leaks and accuracy
  • Response history and before-and-after failure evidence
  • Email alerts with severity, reproduction steps, and fix guidance
  • Monthly health report and remediation Markdown
Explore Watch
For growing AI products

Patrol

$499/mo

Billed monthly

Track security and answer quality across your bots. Catch regressions after prompt, model, retrieval, and vendor changes.

Start Patrol
  • Everything in Watch, across up to 500 bot or AI endpoints
  • Daily prompt injection, jailbreak, and sensitive data probes
  • Weekly multi-turn attacks and retrieval poisoning scenarios
  • 10 canary phrases and 50 business facts across the plan
  • Behavior diffs, brand voice checks, and model regression tracking
  • Bot Health Score across safety, accuracy, leaks, and availability
  • Slack and email alerts; monthly and on-demand reports
Explore Patrol
For agents and MCP

Sentry

$999/mo

Billed monthly

Monitor the tools and permissions behind your AI. Test authenticated agents and MCP servers for changes that expose data or expand access.

Start Sentry
  • Everything in Patrol, with unlimited bot, agent, or MCP endpoints
  • MCP tool inventory, schema changes, and permission drift monitoring
  • Daily adaptive prompt injection, tool and agent poisoning tests
  • Authenticated journeys, role boundaries, and token revocation probes
  • 25 canary phrases and 100 business facts across the plan
  • RAG leakage, memory poisoning, and approved tool misuse scenarios
  • Webhook, Slack, and email alerts with traces and fix tasks
  • Managed CI/CD regression corpus and quarterly security evidence pack
Explore Sentry

We confirm endpoint compatibility, access, and test limits during setup. An endpoint is one bot, AI application, or MCP server. Your model and infrastructure usage charges are separate.

02 / One-time testing

Map the exposure. Prove the risk. Test the limits.

AI agents explore the attack surface and adapt tests as they learn. Our penetration tests add expert verification, business impact, and remediation your team can act on.

Map your exposure

Agentic Surface Scan

$999

One-time engagement

AI agents explore your public attack surface and build a prioritized map of potential weaknesses and areas to investigate.

Get your surface scan
  • Attack surface map of discovered assets and entry points
  • Potential weaknesses ranked by likely impact
  • Evidence and context for each observation
  • Recommended fixes and deeper testing priorities
  • Downloadable report and engineering action list
See coverage and deliverables
Validate your highest risks

Focused Pen Test

$8,999

One-time engagement

A focused penetration test combining agentic exploration, adaptive AI attacks, and expert verification across your most important workflows.

Scope your pen test
  • Everything in the surface scan
  • Focused web, API, AI agent, and MCP penetration testing
  • Verified findings with reproduction evidence and impact
  • Executive summary and prioritized remediation plan
  • OWASP coverage record, CWE classifications, and CVSS severity rationale
  • Security evidence for customer, SOC 2, HIPAA, and PCI DSS reviews
  • Markdown fix tasks for engineers and coding agents
  • Findings walkthrough and one retest within 90 days
See coverage and deliverables
Test the whole system

Full Pen + Load Test

$14,999

One-time engagement

A full penetration test and controlled load test across your agreed application, AI agents, MCP integrations, and infrastructure. Find what breaks and how far it reaches.

Scope your full test
  • Everything in the Focused Pen Test
  • Full penetration test plus controlled load testing
  • Permission and blast radius maps across connected systems
  • Source, cloud, and release pipeline review with provided access
  • Capacity findings, bottlenecks, and scaling recommendations
  • Leadership report, engineering runbook, and fix tasks
  • Framework crosswalk agreed in scope, plus evidence and coverage limitations
  • Findings walkthrough and one retest within 90 days
See coverage and deliverables

Scope, access, test depth, and delivery schedule are agreed before payment. Both penetration tests include a findings walkthrough and one retest of the original findings within 90 days.

Choose the depth your product needs
CoverageAgentic Surface Scan$999Focused Pen Test$8,999Full Pen + Load Test$14,999
Public attack surfaceDiscovery and risk mapIncludedIncluded
Authenticated web and API testsPublic surfaces onlyPriority workflowsFull agreed scope
AI agents, RAG, and MCPPublic exposure signalsTargeted attack scenariosAttack chains and poisoning
Identity and permissionsPublic signalsRole and tenant checksScopes, revocation, and blast radius
Source, cloud, and supply chainPublic exposure signalsRelevant external exposureReview with provided access
Load and resilienceNot includedNot includedControlled load test included
ReportPotential weaknesses and next stepsVerified findings and fix tasksFindings, capacity, and engineering runbook
Walkthrough and retestNot includedOne retest within 90 daysOne retest within 90 days
Built for the next decision

Give engineering a fix list.
Give leadership the facts.

Pen test reports connect technical evidence to the customers, permissions, and systems at risk. The full engagement adds blast radius maps and measured load results.

Inspect the sample report
01

Reproduce the finding

Affected flow, attacker preconditions, observed result, and redacted evidence.

02

Understand the reach

Which identities, tools, customer data, and connected systems are exposed.

03

Ship and verify the fix

Prioritized tasks, coding agent instructions, validation checks, and retest criteria.

Start with a free assessment

Get a first read on your risk.

A high-level look at your public exposure and the testing that would be useful next. Step up to the $999 scan for a detailed map and investigation plan.

Get a free assessment
Before you start

A few practical details.

Yes. Pen test reports provide scope, methods, verified findings, OWASP coverage, severity rationale, remediation, and retest evidence for SOC 2, HIPAA, PCI DSS, ISO 27001, and NIST-aligned reviews. Tell us which framework and control mappings your reviewers need during scoping. A security assessment supports that work; it does not itself provide certification, SOC 2 attestation, a HIPAA compliance determination, or a PCI ASV scan.