Encryption, identity, access, exposure, AI.
The secure option at every decision.
AI assistants can ship services on plain HTTP, leave admin consoles and AI tools on default credentials, and keep default settings unless told otherwise. SecureConfig gives your assistant short, copy-ready guides that make the secure option the default at each deployment decision. Guides include checks your assistant runs and reports, including what it could not verify.
Browse the guides
Start here
- Browse the guides102 guides102 guides in 12 categories, from certificates to AI infrastructure, each read on GitHub or fetched raw.
- Point your assistant at the rules10 rulesThe 10 rules an assistant applies to every service it configures, and the raw URL to fetch each guide.
- Copy a ready-made task6 tasks6 prompts to paste into your assistant: secure the project, review the AI stack, add single sign-on, and more.
- Run the verification checklist11 checks11 checks to run before reporting completion, with the command or method for each.
What it gives you
An AI assistant that sets up a service does not always close every open door unless it is told to: encryption left off, default credentials for logins, factory settings kept. SecureConfig gives it a guide for each of those decisions, drawn from the vendors' documentation and free to use under CC0, so the secure setting goes in the first time. Guides include checks that report what passed, what failed, and what could not be verified. The checklist's steps are reasoned from documented vendor behaviour, with no deployment run recorded, and most guides record the versions they were written against.
About this project