Know what your AI tools touch when the work is confidential
You use AI on real client work: drafting, research, review. Every assistant and MCP tool can read your files and reach the network, and you have no record of what it touched. Provae for AI shows you, on your own machine. It collects nothing, and you can verify that yourself.
Get early accessThe problem
You adopted AI faster than anyone gave you a way to see what it does. The assistant, the agent in your editor, the MCP servers you installed because they are useful: each one can read files and reach the network, and none of them shows you what it actually touched.
An agent with a filesystem tool and a fetch tool can read a privileged document and send it anywhere, and the first you would hear of a problem is after it happened. You would not run a background process with that much access and no logging. For most people doing confidential work with AI, that is the current setup.
What Provae for AI does
- Watches every AI tool call on your machine. It sits between your AI client and its tools and records what ran, which files were touched, and whether anything sensitive was involved, including a secret or a client detail pasted straight into a prompt. Locally. Nothing leaves.
- Flags tampering. If a tool you approved quietly changes what it does, you get an alert. This is the most common way an AI agent turns against you.
- Keeps a record for your own diligence. A local log, signed so you can confirm it has not been altered: what ran, which categories of sensitive data were touched, and whether anything was tampered with. It is your record, for your own confidence and risk management. Not a certificate you hand to anyone else.
What it shows, and what it doesn't
We will not overstate this, because a tool that oversells what it sees is worse than no tool.
It shows you, on your machine, what your AI tools touched and whether any were tampered with. It does not yet show where data ultimately went across the network; that is the layer we are building next. Today it answers what your AI tools accessed and whether they are still what you approved. It does not yet answer whether anything left, and to where.
And it is a record for your own diligence, not a compliance certification or a guarantee for a regulator, bar, or insurer to lean on. You keep it, and you verify it yourself.
Zero collection, and you can check
Provae for AI stores everything in a local file you own. It has no server to send your activity to. The sensitive details it inspects, it reads in memory and never writes down; the record holds categories and counts, never your file paths or their contents. You do not have to trust that. The tool is built so you can confirm it.
Who it's for
Solo and small-practice professionals who handle information they cannot afford to leak and now use AI on it: lawyers, accountants, therapists, consultants. And any developer or power user running agents on a machine that holds sensitive work. If your AI tools touch things that matter, this shows you what they are doing.
We store the email you enter and nothing else: no analytics, no tracking, no list-sharing. One email when it is ready; reply "remove" any time and it is deleted. Prefer your own mail client? Email us instead.