Our small research team recently started using AWS, and resources are being created and managed by different people without a consistent inventory. We are considering infrastructure as code, but first need a reliable way to discover what exists in our account, understand resource relationships and configuration, and possibly track changes over time.
Because our projects change frequently, we regularly encounter unfamiliar AWS services and workflows. We would like to generate an inventory in a structured format such as JSON, CSV, or SQL results that can be used as context for an AI assistant and also help us write or generate infrastructure-as-code definitions. We are trying to avoid building a discovery system from scratch. What established tools or AWS features would be good starting points, and what limitations should we be aware of?
4 Answers
Another practical approach is to use a read-only AWS integration with an AI coding assistant, then ask it to inventory resources, explain the architecture, and produce documentation or IaC drafts. Keep the permissions strictly read-only and have the output reviewed by a person before applying anything. Native discovery tools are generally better for collecting complete data, while the AI assistant is more useful for organizing it, explaining unfamiliar services, and turning selected resources into maintainable code.
Former2 can scan an existing AWS environment and generate CloudFormation templates, so it may be useful for turning parts of an undocumented setup into an initial infrastructure-as-code baseline. I would treat the generated templates as a starting point rather than production-ready code: review dependencies, secrets, permissions, unsupported resources, and environment-specific values carefully.
AWS Config is probably the best native starting point. It provides an inventory of supported resources, includes relationships between them, and can retain historical configuration changes. It is useful for more than just a one-time export, since you can also use it for compliance checks and remediation. For multiple accounts, configuration data can be aggregated centrally. For a single account, the Tag Editor may also provide a quick, though less complete, overview.
Steampipe is a strong option if you want something easy to query and export. It exposes AWS resources through SQL, so you can build focused inventory queries and write the results to JSON or CSV for documentation, automation, or AI context. It also lets you use narrowly scoped read-only permissions. Be mindful that broad queries across a large account can generate a lot of API calls, so start with the services and regions you actually need.

Related Questions
Can't Load PhpMyadmin On After Server Update
Redirect www to non-www in Apache Conf
How To Check If Your SSL Cert Is SHA 1
Windows TrackPad Gestures