partner program

Recommend the tool that stops rm -rf.
Earn 30% — for life.

You teach AI coding, or your newsletter reaches developers who run agents every day. gate.cat is the deterministic deny-list that vetoes an agent's catastrophic shell commands before they run. Share your link, and every paid plan someone starts from it pays you 30% of every payment — not just the first, for the whole lifetime of the subscription.

Get my partner link → Watch the 30-second veto demo →

No cost, no minimums, no exclusivity. The blocking core is free and open source (Apache 2.0) — so you're recommending a tool your audience can actually use, not a paywall.

Why this converts for your audience

Every developer who runs Claude Code, Cursor, Codex or aider has handed a shell to a model. The failure mode is not hypothetical — it is the single thing your audience is quietly afraid of.

Real fear, real fix.

A Cursor agent wiped a production database and every backup in ~9 seconds. Another burned ~$47k in a runaway loop. gate.cat is the deterministic layer that stops exactly those, before they run — an easy, honest recommendation.

Free core = trust.

You send people to something they can install and use for €0. Nobody feels sold to. The paid Cloud/Team layer (off-machine record, fleet alerts) is what earns you commission — and only teams tend to want it.

Recurring, not one-shot.

Attribution is keyed to the subscription, so renewals keep paying you — for as long as they stay subscribed. A single good video or newsletter mention can pay out for months.

A demo that sells itself.

Link the 30-second veto demo. Your audience watches an agent get stopped mid-rm -rf. The product does the convincing; you just point at it.

The terms — plainly

30%

of every payment

Solo, Team (€149/mo flat), Business (€399/mo) and one-time policy packs — all pay 30%.

for life

recurring, not first-order

Every renewal on a subscription you referred keeps paying, for as long as it stays active.

auto

tracked & accrued

Your ?ref=code sets a 90-day cookie and rides the checkout automatically. The ledger accrues your commission; refunds claw back, so the numbers stay honest.

Payout by PayPal or bank transfer. Free core is never behind the affiliate wall — you always send people to real value first.

How it works — three steps

Step 1

Get your link

Email for a code and you get gate.cat/?ref=yourname. It works on any page — the homepage, /teams.html, or /veto-demo.html?ref=yourname.

Step 2

Share it honestly

Mention it where it fits — a Claude Code tutorial, a "how I stopped my agent from nuking my repo" post, a DevTools newsletter slot. Point at the free core and the demo.

Step 3

Get paid on every payment

Anyone who upgrades to a paid plan from your link earns you 30% — on the first payment and every renewal after. We reconcile and pay out on the ledger.

What you're pointing people at

Numbers already published in the repo — so when your audience checks, the claims hold up. That's the whole point of a trust-based recommendation.

178 / 178 dangerous commands blocked in the threat suite

1,085,159 real agent commands replayed

1 documented false-block (published in the repo)

Recommend it honestly — what it does not do

Keep it straight with your audience and it stays a trust win. gate.cat catches known dangerous command shapes with a deterministic deny-list. A deliberately obfuscated or genuinely novel command can still slip past it — the full bypass suite is in the repo. It is not a sandbox, not an antivirus, and not a substitute for least-privilege credentials. Please don't call it "compliant", "certified" or a "guarantee" — it's the fast deterministic layer that stops the well-known catastrophic mistakes an AI agent makes far more often than a human would, before they run. Honest framing is why people buy.

Try it yourself first — the core is the whole engine.
pip install gate.cat, wire it into your agent, watch it veto rm -rf. Once you've seen it work, recommending it is easy. Then email for your partner link.