APPLICATION SECURITY REVIEW

Review your app.
Get a written fix list.

An independent review of your code, architecture, and AI features with Keith Myers. Get findings ranked by severity, with a recommended action for each.

  • Authentication, secrets, data protection, and AI risks
  • Reviews for established applications and apps built with AI coding tools
  • Scope, access, and fee agreed before work begins

30 minutes · No charge · Opens CalendarBridge

Keith Myers, founder of Entori
Keith Myers25 years in IT and security · CISSP · CISMAbout Keith

DISCUSS YOUR REVIEW

Tell me about your app.

Share what the app does, the data it handles, and what you want reviewed. Keith usually replies within one business day.

Your message goes directly to Keith. Please leave passwords and sensitive security details out of this form.

WHO IT’S FOR

Before your app meets real users and real data.

  • Teams shipping apps built quickly with AI coding tools such as Cursor, Claude Code, Lovable, or Replit.
  • Software companies preparing for a client security review or due diligence.
  • Products that handle personal, financial, or regulated data.
  • AI features built on models, vector databases, agents, or third-party integrations.

WHAT I REVIEW

Six domains.
One consistent method.

01

Identity & access

Authentication on every endpoint, least privilege for integrations and service accounts, secrets handling, and vendor access to production.

02

Configuration & deployment

Production servers, exposed ports, network boundaries, container settings, and environment configuration.

03

Encryption

Encryption in transit between every service, encryption at rest, and field-level protection for sensitive data.

04

Data protection

Personal and regulated data across databases, vector stores, backups, and snapshots, with classification and retention.

05

Logging & audit

Centralized logging, database audit, and a tamper-resistant record of what the system and its AI components did.

06

AI-specific risks

Prompt injection, model tool permissions, data sent to model providers, and the human review around AI output.

02 / THE WORK YOU KEEP

A findings report
your engineers can use.

Each finding has a domain, a severity, and a recommended action. Critical items are the ones to close before production.

Critical

Directly exploitable today, or a regulatory blocker. Close before production.

High

A meaningful gap in defense, data protection, or compliance. Address promptly.

Medium

A strong recommendation for production readiness or best practice.

Low

Minor configuration or housekeeping. Fix in the normal development cycle.

Open

A question that needs an answer before the finding can be rated.

Illustrative excerpt of a findings register
#DomainFindingSeverityRecommended action
01IAMAPI endpoints accept requests with no authentication. Network location is treated as trust.CriticalAdd authentication on every endpoint, deny by default, and log the user on each request.
02IAMSecrets and credentials committed to source control.CriticalMove secrets to a managed secret store, rotate every value, and remove them from the repository.
03ConfigDevelopment web server running in production.MediumReplace with a production-grade application server.
04DataCustomer identifiers embedded into the vector database without redaction.HighRedact or tokenize sensitive values before embedding. Treat embeddings as personal data.
05AIUntrusted user content can steer the model’s tool calls.HighConstrain tools to named, parameterized operations and allow-listed data.
06OpenUnable to confirm the main branch matches what is deployed.OpenConfirm where production code lives before the findings are finalized.

Illustrative sample based on common findings. No client details. A redacted sample report is available on request.

HOW IT WORKS

From repository access
to a clear fix list.

01

Agree the scope

A free 30-minute intro call to understand the application, its users, and its data. Agree access and scope.

02

Review the application

Read-only review of your repositories, architecture diagram, and documentation, plus a working session with your team.

03

Deliver the findings

A written report with every finding rated by severity, the assumptions behind it, and a recommended action.

04

Walk through the fixes

Review priorities with your engineers. Close the critical items first, then work through the rest.

A FEW USEFUL ANSWERS

Before we talk.

No. This is a review of your code, configuration, architecture, and documentation. If a penetration test is warranted, the report will recommend it.

DISCUSS YOUR APPLICATION

Agree the review scope
with Keith.

A free 30-minute call to discuss your app, its data, and the review.

Book a free intro call Or send an inquiry