Engineering support for StackQL
StackQL is an open-source tool for querying, provisioning and operating cloud and SaaS resources with SQL. Dignep has provided engineering support to the StackQL team.
The client
StackQL is an open-source project from StackQL Studios, whose GitHub organisation lists Australia as its home. Its own description is an infrastructure-as-code tool that lets people and AI agents work with cloud control planes and data planes using SQL.
The core application is written in Go and released under the MIT licence. It parses SQL statements, turns them into API requests to the provider, runs them and returns the results as rows.
This StackQL engineering case study describes the project and the role we've played. It doesn't claim credit for the product, which belongs to the StackQL team.
What the project covers
StackQL has a wide surface for an open-source project. These are the main parts, as published on its site and GitHub.
SQL engine
A standalone Go application used through exec or shell, or in server mode over the Postgres wire protocol, so standard Postgres clients can connect.
Provider registry
Provider interfaces for cloud and SaaS platforms such as Google, AWS, Azure, Okta and GitHub, defined as extensions to each provider's OpenAPI specification.
MCP server
StackQL runs as a Model Context Protocol server, so MCP-capable AI clients can query and provision resources across more than 40 providers using SQL.
stackql-deploy
A declarative infrastructure-as-code framework that works without a state file. The current version is written in Rust.
Documentation site
Getting-started guides, command-line usage, a full language specification, MCP docs, tutorials and how-to guides written for AI agents.
Distribution
Packages for macOS, Windows and Linux, Docker images and GitHub Actions. The MCP server is also published to npm, PyPI and the official MCP Registry.
Our role: engineering support
Dignep has worked with the StackQL team in an engineering support role. We don't publish the specifics of that work on this page. The project is open source, so its code, issues and release history are public on GitHub for anyone who wants to see how it is built.
For our own engineers, a project like this is useful practice in the areas we sell. Cloud provider APIs, infrastructure as code and agent tooling such as MCP all come up in client work on generative AI and agent systems.
What we can help with on open-source projects. Extra engineering capacity that works inside your repository, your review process and your release cadence, without taking over the roadmap.
Stack
Taken from the project's README, its GitHub repositories and the live site's HTML and headers.
| Layer | What the project uses |
|---|---|
| Core | Go, MIT licence |
| Interfaces | Interactive shell, exec, Postgres wire-protocol server mode, MCP server |
| Provider definitions | OpenAPI 3 extensions, stored in the StackQL provider registry |
| Deployment tool | stackql-deploy, written in Rust |
| Website | Docusaurus 3, hosted on Netlify |
Have a project like this?
Book a 30-minute discovery call. Tell us about your codebase and where you need extra engineering hands, and we'll send a written proposal within two working days.
