Control what your agents send, and prove it.
H-Gate inspects the data an agent is about to send across a boundary and applies your policy: allow, redact, or block. Every redact or block is anchored as an H-Seal receipt. Free here, $0.10 per inspection on the agent API.
How it works
H-Gate sits at the chokepoint where an agent sends data: an outbound API call, an agent-to-agent message, a context fetch.
A deterministic engine flags emails, cards, government ids, financial accounts, crypto addresses, provider tokens, phone numbers, secrets, keys, and PHI. Pattern detectors plus checksum, entropy, and Luhn validation, no model in the trust path.
Allow, redact, or block by entity type and sensitivity. The decision is the strongest per-finding action; redacted text masks every match.
Every redact or block on the paid agent path anchors an H-Seal receipt (hashes and counts only, never the plaintext). The detector is enclave-safe by design: raw matched values are never returned or logged, only hashes, types, and spans.
What it catches
A deterministic detector suite for the data that should not leave without a decision. The engine is pure and enclave-safe by design (no I/O, no clock, no network; only hashes, types, and spans leave it); the catalog is extensible.
- US SSN
- Phone (E.164 and US)
- IPv4 address
- IPv6 address
- PHI markers
- Credit card (Luhn-validated)
- IBAN (mod-97)
- US ABA routing (checksum)
- SWIFT / BIC
- Ethereum (EIP-55)
- Bitcoin
- XRP Ledger
- Stellar
- Hedera
- Solana
- API keys and tokens
- AWS access keys
- Provider tokens (Stripe, GitHub, Google, Slack, OpenAI, Anthropic, Twilio, SendGrid, npm)
- PEM private keys
- JWTs
- High-entropy secrets
Pricing
Inspect any payload here at no cost. Full engine, full policy set.
Paid via x402: USDC on Hedera, Base, Solana, Stellar, or RLUSD on the XRP Ledger. Every redact or block anchored as an H-Seal receipt.