agentprime
  • Ex-Datadog security leadership
  • Fortune 500 security programs
  • Fixed-price engagements

Services

Security Assessment

A fixed-scope review of the product you already run, from the code and cloud to who has access, with findings ranked by risk and a plan to fix them.

What's included

  • Threat model of your product and architecture, covering where sensitive data lives and who can reach what
  • Application and code review aimed at the mistakes that matter most, like missing authorization checks, one customer seeing another's data, unchecked input and secrets in code, with special attention to code written by AI tools
  • Cloud and infrastructure configuration review across hosting, databases, DNS and every connected service
  • Identity and access review covering SSO, 2FA, admin rights, offboarding, developer laptops, and the AI tools and connectors on them
  • Spend and abuse limits, logging and alerting, so you'd know if something went wrong
  • A security check wired into your deploy pipeline, tuned to your stack
  • Findings ranked by risk, with fixes your team or your AI tools can apply
  • A 90-day roadmap, a written report, a one-page summary for customers and your board, a walkthrough and one retest

Who this is for

Your product is live and growing. Customers depend on it, and the bigger ones have started asking how you secure it. It was built fast, by a small team, with AI tools doing more and more of the typing. Nobody has ever looked at the whole thing with an attacker’s eye.

That’s what the assessment is for: an honest, senior look at what you actually run today, and a clear plan for what to fix first.

What we look at

We start where small companies most often get hurt:

  • Authorization gaps. Log in as customer A, change an ID, see customer B’s data. The most common serious bug we find, and AI coding tools make it more common.
  • Keys and access with too much power. Full-access keys where restricted ones would do, former contractors who still have access, admin rights nobody remembers granting.
  • No limits and no trail. Nothing stops abuse of paid APIs, and when something odd happens there are no logs to say what.
  • Your AI tools. Agents and editors that run with a developer’s full permissions, and connectors that reach production data.

Then we work outward through the rest of the system, scoped in writing before we start.

What you’re left with

Not just a report. You get a ranked fix list, a roadmap your team can follow without us, a one-page summary you can share with a customer or your board, and a security check running in your deploy pipeline so new code gets reviewed before it ships.

Launching something new?

The same assessment works before a launch. For a single new app, it usually takes about a week.

How access works

We use each service’s own read-only access wherever we can. We never ask for passwords or API keys by email, chat or document. We don’t change your production systems without your written approval.

Not sure where you stand? Let's find out.

Tell us what you run and what's coming up: a big customer, a security questionnaire, an insurance renewal. We'll tell you plainly whether we can help and what it would take.

You work directly with Craig and Justin.No account managers, no junior hand-offs.

Talk with a founder →