# OneHuman

> Self-hosted controls for AI agents using customer accounts: mask private fields, require passkey approval for sensitive actions, and record signed decisions.

OneHuman helps application teams control AI agents acting inside signed-in customer sessions. Its browser SDK and server middleware assess session signals, apply per-route policies, mask selected private fields and require human passkey approval for sensitive actions. Signed decision records support later review, and observe mode lets teams evaluate policies before enforcement. It is designed for customer-agent access control, rather than workforce AI monitoring or general perimeter bot filtering. The SDK and integrations use Apache-2.0; the engine uses Business Source License 1.1 with an additional production-use grant and restrictions on competing hosted or embedded offerings. Detection is not infallible, so teams should test their policies and review the documented limitations.

## Quick facts

- Categories: Developer Tools, Security and Privacy
- Pricing: Freemium (free plan + paid tiers) — Official pricing checked 2026-10-07: Free includes up to 100 agent-active customers per month; Pay as you go is $0.50 per additional customer; Growth is $299/month including 1,000, then $0.30 each. Enterprise pricing is custom. A portal API key is required for protection on a real domain; without a key, the documented default is localhost development only. Confirm current terms and limits with the provider.
- Free plan: Yes
- Official site: https://onehuman.ai/
- Last checked: 2026-10-23

## Key features

- Per-route controls for customer-operated AI agents
- Masking of selected private response fields
- Passkey approval for sensitive actions
- Signed decision records with offline verification
- Observe mode before enforcement
- Self-hosted engine with browser SDK and server integrations

## Use cases

- Protecting customer data in signed-in SaaS sessions
- Requiring approval before exports or sensitive account actions
- Evaluating customer-agent activity before enforcing policies
- Reviewing signed access-control decisions

## Links

- Nextool.ai page: https://nextool.ai/tools/onehuman/
- Alternatives: https://nextool.ai/alternatives/onehuman/

Source: Nextool.ai (https://nextool.ai/tools/onehuman/). Pricing and features change; confirm on the provider's site before relying on them.