Pricing
Simple pricing. Serious protection.
Start free. Pay once when you need the full arsenal.
Free
no credit card, ever
- Basic policy engine (ALLOW / APPROVE / BLOCK)
- Basic shield — original 12 prompt-injection patterns
- Kill-switch + anomaly alerts
Pro
one-time — yours forever
Just +$79 one-time over Free
Your $79 covers the full ~70-pattern scan engine, policy engine and red-team suite — servers, updates and support.
- Everything in Free, plus:
- Infection Chain Map — patient-zero tracing + one-click chain quarantine
- Shield v2 — ~70 advanced + novel patterns, stego detection, decode + rescan, homoglyph normalization
+ 6 more in the arsenal below ↓
Already have a key? Redeem it →
Bitcoin checkout — one-time payment, same never-expiring license key as card checkout. No auto-renewal, no subscription.
Team
one-time — yours forever
Just +$70 one-time over Pro
- Everything in Pro, plus:
- Autopilot auto-approve — high-confidence, low-risk repeats approved inline, fully audited
- 5 seats
Already have a key? Redeem it →
Bitcoin checkout — one-time payment, same never-expiring license key as card checkout. No auto-renewal, no subscription.
Every plan covers every agent in one workspace. No per-agent fees, no seat math.
Card checkout is being restored. Use the WhatsApp or email buttons on this page and we handle your purchase personally today.
Once we confirm your payment, we email your license key to you.
The key lands in your inbox — paste it on the billing page and Pro is live.
Card checkout is being restored. Buy direct today: message us, we confirm payment and email your license key — redeem it on the billing page.
💬 Buy Pro on WhatsApp — $79 one-time 💬 Buy Team on WhatsApp — $149 one-time
✉️ Buy Pro by email ✉️ Buy Team by email
Billing questions: [email protected]
Redeem key
One-click refund, no questions asked · Pay once — yours forever, no subscription · We reply personally.
Pro-exclusive arsenal
What your $79 unlocks
The features below are Pro-gated server-side — Free workspaces get a 402 upgrade gate, never a watered-down version. This is the value gap.
| Feature | Tier | Why it matters |
|---|---|---|
| Infection Chain Map | Pro | Traces viral prompts agent→agent back to patient zero; quarantine the whole chain in one click. |
| Chain quarantine | Pro | One button quarantines every infected agent in the chain and fires an alert. |
| Autopilot recommendations | Pro | Every approval carries your team's learned recommendation + confidence + evidence. |
| Autopilot auto-approve | Team | High-confidence (≥0.90), low-risk repeats approved inline — humans audit everything. |
| Sleeper Shield | Pro | Catches time-bomb instructions and extracts when they're armed to fire. |
| Shield v2 | Pro | ~70 attack patterns: stego, homoglyph normalization, decode + rescan, multilingual bypasses. |
| Policy v2 | Pro | Velocity limits, cost anomalies, daily caps, first-seen tool, destructive-action review. |
| Canary tokens + exfil guard | Pro | Planted fake secrets — a leak proves your agent would leak real ones. |
| Red-team suite | Pro | 55-attack benchmark: quantified proof for clients — "our shield blocks X% of N techniques". |
| Behavioral baselines | Pro | Per-agent behavioral baselines feed the anomaly detection on every plan's kill-switch. |
| Memory Write Guard | Pro | Your agent stops learning the attacker's instructions as facts it will obey next week. |
| Spawn Tree Guard | Pro Team pause | One runaway task can no longer quietly turn into 800 paid tasks and a five-figure bill. |
| Mission Drift Detector | Pro | You find out your agent left the job you gave it before it finishes a different job you never approved. |
| Canary Tools | Pro | The first time an attacker steers your agent, it reaches for a tool that proves the attack and stops itself. |
| Tool Argument Injection Firewall | Pro | An allowed tool stops being a clean way to smuggle a banned command into your systems. |
| Context Bomb Guard | Pro | A poisoned data source can no longer bankrupt a run or bury the real instruction by flooding your agent's memory. |
| Approval Forgery and Replay Detector | Pro Team report | Your agent can no longer talk its way past the one control you trusted by claiming you already said yes. |
| Agent Collusion Channel Detector | Team | Two compromised agents can no longer quietly coordinate through your own workflows while each one looks clean alone. |
| Model Swap Re-Attestation Gate | Pro Team hold | A model change you never chose can no longer quietly change what your agent will do with your money and data. |
| Adversarial Cost-Imposition Shield | Pro | A stranger can no longer run up your AI bill on purpose by feeding your agent work that never finishes. |
| Before-Image Rollback | Team | When an agent goes rogue, you undo what it did instead of cleaning it up by hand for days. |
| Stale Instruction Replay Guard | Pro | Yesterday's instruction can no longer come back as today's order after you changed your mind or your rules. |
Team also adds shared workspaces and priority alert review on top of Autopilot auto-approve and 5 seats.
Compare every feature
Every row is enforced server-side — Free workspaces get a 402 upgrade gate on Pro rows, never a watered-down version. Plans are one-time payments; there is no annual billing because there is no subscription to bill.
| Feature | Free | Pro — $79 once | Team — $149 once |
|---|---|---|---|
| Policy engine (ALLOW / APPROVE / BLOCK) | Basic | v2 — velocity limits, cost anomalies, daily caps, first-seen tool, destructive-action review | Everything in Pro |
| Injection shield | 12 patterns | Shield v2 — ~70 patterns, stego detection, decode + rescan, homoglyph normalization | Everything in Pro |
| Kill switch + anomaly alerts | Included | Included | Included |
| Infection Chain Map + chain quarantine | — | Included | Included |
| Sleeper Shield (deferred-instruction detection) | — | Included | Included |
| Canary tokens + exfiltration guard | — | Included | Included |
| Red-team suite (55-attack benchmark) | — | Included | Included |
| Behavioral baselines per agent | — | Included | Included |
| Autopilot recommendations | — | Included | Included |
| Autopilot auto-approve | — | — | Included |
| Seats | 1 workspace | 1 workspace | 5 seats + shared workspaces |
| Agents covered | Every agent in the workspace | Every agent in the workspace | Every agent in the workspace |
Pricing questions
How much does Rogue-Agent Defense cost? The Free plan is $0, Pro is $79 one-time, and Team is $149 one-time. There are no subscriptions and no per-agent fees — every plan covers every agent in one workspace. You buy a license key once by WhatsApp or email, redeem it on the billing page, and the tier is yours forever.
What if Rogue-Agent Defense doesn't fit my setup?
One-click refund, no questions asked. Email [email protected] and your license-key purchase is refunded in full — the key is revoked and the money goes back to you.
What does "one-time" mean?
You pay once and the tier is yours forever — no subscription, no renewal, no surprise charges. The Free tier stays $0/mo with no credit card, ever.
How many agents does each plan cover?
Every plan covers every agent in one workspace — no per-agent fees, no seat math on Free or Pro. Team adds 5 seats and shared workspaces.
What happens after I buy?
You get a license key by email once we confirm your payment. Paste it on the Billing page and your workspace unlocks instantly — shield v2, policy v2, canaries and the red-team suite.
Rogue-Agent Defense Pro — $79 one-time: full injection shield, policy engine, canaries.