Kyvvu

Quickstart

Run your first agent under enforcement.

You can try the engine on your own machine without an account, and connect the platform when you want the dashboard.

Open the playground

Step 1

Set it up in a few lines.

Pick the tab that matches what your agent runs on. The same engine runs in each of them.

In your process
No proxy and no extra hop. Prompts and data stay on your machine.
Under a millisecond
Policies are fetched once and cached. Every verdict is computed locally.
Policies in your repo
YAML manifests next to your code, reviewed like any other change.
$ uv tool install "kyvvu-cli[claude-code]"# Connects to the API, registers the agent, installs the hooks$ kyvvu claude setup

Step 2

Start in shadow mode, turn on blocking when you're ready.

Turn enforcement on one agent at a time, with the same policy file all the way through.

  1. Shadow

    Watch what agents do

    Kyvvu records every verdict and blocks nothing. You see which rules would fire, on real traffic.

    step.message POST smtp/external, after a PII read

    ALLOWruns · logged as a would-be block

  2. Suggest

    Flag before you block

    Risky steps still run but get a WARN, so owners can tune policies without stopping work.

    step.message POST smtp/external, after a PII read

    WARNruns · flagged to the owner

  3. Enforce

    Stop what shouldn't run

    Dangerous sequences are blocked before they execute, and every verdict lands in your SIEM.

    step.message POST smtp/external, after a PII read

    BLOCKstopped before it ran

Next

Go further.

In your browser

The playground

Build an agent's path and watch the engine decide, step by step.

With your own logs

Trace check

Drop a log of what your agent did and see which standard policies would have stepped in.

The reference

Developer docs

The SDK, every integration and the policy rules reference.