Is your agent acting like you would?
Thunzi reads what your AI agents do on your Mac and tells you when one crosses your line. Your data never leaves your computer.
For Mac. Works with Claude Code and Codex. Reads logs only. Changes nothing.
You asked: "fix the CI deploy step." You did not ask for this.
How it works
Your agents send email, push code and spend money for you. Each tool has its own permission prompts, and most of us click through them. You can't read seven thousand tool calls a week. What your agents do still reflects on you.
Reads your agents' logs
Claude Code and Codex already write down what they do on your Mac. Thunzi reads those logs. Nothing to reroute or wrap.
Checks them against your line
Clear rules catch risky actions, and Thunzi checks whether you actually asked for them.
Tells you in one line
Urgent things arrive as a notification. The rest lands in a short daily digest, grouped so forty retries become one card.
Thunzi starts by reading your last 30 days, so you see results on day one.
Where agents most often cross the line
Four places to start. Most come straight from our own testing.
~/.claude/…/6f1c.jsonl:412
Security
A live GitHub token printed into a transcript. A repo created public. A token parked in /tmp.
vercel deploy --prod .env.local
Leaks
Secrets or private data headed for a public deploy, a Slack post or a shared file.
to: dana@clinet.co (guessed)
Wrong place
Emails sent to a guessed address. Files written outside the project.
credits: no cap set
Spend
A purchase made around a block. Credits burned with no cap.
Coming next: your own rules. Write a few values in plain words, like “never email a customer without me” or “keep client work out of public repos.” Thunzi flags actions that go against them.
One line, with a fix ready.
An agent printed a live GitHub token
~/projects/app · Today, 11:22 AM PT · 3 events
git credential fill returned ghp_****Xy9. It is now in plaintext in the session log, line 412.
You asked: “fix the CI deploy step.” You did not ask for this.
An agent emailed a customer without you
Your rule: “Never email a customer without me.”
To: dana@client.co · Oct 3, 4:08 PM PT
Every alert answers five questions: what happened, where, when, did you ask, and what to do.
Your agents, your Mac.
How Thunzi is built:
- Logs are read, never changed.
- Secrets are masked and hashed before Thunzi stores anything.
- Nothing is sent off your Mac.
- No content telemetry. No open ports.
Everything stays on your Mac.
Why Thunzi
Isithunzi is the isiZulu word for shadow. It also means dignity: the standing and respect a person carries, which their actions can build up or wear down. Its root, umthunzi, is shade, the cool cover of a tree. Thunzi is your shadow on every agent, keeping what they do in line with your values and covering you without getting in the way.
Say it TOON-zee.
We ran it on our own agents first.
Before building the app, we replayed one week of one founder's agents (7,466 tool calls, 16 agents) through Thunzi's rules.
FAQ
Who is Thunzi for?+
Anyone running AI agents on their own Mac: founders, developers, solo operators, power users.
Which agents does it work with?+
Claude Code and Codex.
Does it block actions?+
Not at first. The first version watches and tells you. Stopping risky actions before they run comes later, through the hooks these agents now support.
Does my data leave my Mac?+
No. Thunzi reads and stores everything on your Mac.
Can I set my own rules?+
Soon. Thunzi starts with sensible rules for the four areas above. Next, you'll write your own values in plain words and Thunzi will hold your agents to them.
What does the name mean?+
Isithunzi is isiZulu for shadow, and also for dignity: the standing and respect a person carries. Umthunzi is shade, the cover of a tree. Thunzi is your shadow on every agent, holding it to your values while staying out of your way.
How do I say it?+
TOON-zee. The th is a soft t, as in “top.”
When can I use it?+
We're letting in a small group of early users. Join the waitlist.