2
1 Comment

Launched Shieldly — AI-Powered AWS IAM security analysis (demo-first, no signup)

IAM security is one of the hardest problems in AWS. I spent years reviewing IAM policies and watching teams make the same mistakes — wildcards in Action and Resource, unscoped trust policies, PassRole chains that nobody had audited since the role was created.

Every tool I found either required an enterprise contract, needed an agent installed before showing a single finding, or was so noisy it was ignored. So I built Shieldly.

What it does:
- Analyzes IAM policies, CloudFormation templates, and AWS costs using AI
- Flags over-privileged roles, privilege escalation paths, wildcard policy risks, and trust policy misconfigurations
- Explains each finding in plain English with a prioritized remediation step
- Ships as web app + CLI + VS Code extension + GitHub Action + CDK Guard (@shieldly/cdk-guard on npm)

The design decision I'm most proud of:
The free IAM Advisor demo requires no signup, no credit card, and no AWS credentials. Paste a policy, get full analysis. I made it this way because the people who will catch policy bugs are the engineers writing them — not the security team that gets assigned a $100k enterprise license. The tool has to be in the hands of the person with the keyboard.

Stack: Next.js, AWS Lambda, API Gateway, DynamoDB, AWS CDK. AI runs on Gemini and Claude Haiku. Auth via Clerk, payments via Polar.sh.

Launch offer: Code 90Off2M — 90% off for the first 2 months. Pro starts at $1.90/mo. Free tier has no paywall at all — just shieldly.io/app/iam.

What I'm working on next: LinkedIn presence, newsletter placements, Product Hunt launch.

Would love feedback on the pricing tiers and the demo experience. What's missing?

— Jeff, jeff@shieldly.io

posted toAvatar for product Shieldly
Shieldly
  1. 1

    One thing I find interesting about tools like this is that they quietly decide who gets to own a problem.

    Some products make security a specialist function.

    Others try to move parts of it back toward the people creating the risk in the first place.

    That choice ends up shaping a lot more than the feature set.