Orivael Intent Firewall

Block harm, deception, and PII leakage in your LLM calls.

A single API in front of any LLM. Returns allow or block in under 50 ms. Every verdict is HMAC-signed for audit.

Start free Quickstart →

One call. Two lines.

curl -X POST https://firewall.orivael.dev/v1/guard/check \
  -H "Authorization: Bearer $AXIOM_KEY" \
  -d '{"text": "What is the weather today?"}'

# {"verdict": "allow", "intent": {"class": "INFORM", ...}}

Or run the firewall from your editor — MCP tools for dev & enterprise.

Axiom ships an MCP server. Drop one config block into Claude Desktop, Claude Code, Cursor, or any JSON-RPC stdio client and get signed tools native to your editor — no extra service to run, no extra key to manage. Core is the free developer set; Enterprise adds the governance, orchestration, and defense tools.

Core

For developers — free tier.

  • axiom_guard_check — constitutional boundary check on any input
  • axiom_lint — lint an .axiom spec for authorship issues
  • axiom_trace — full 3-phase reasoning trace with monotonicity gate
  • axiom_validate.axiom language validator (strict + lenient)
  • axiom_memory — local-first recall over signed memory
  • axiom_axm — AXM container runtime (inspect / verify / run)
  • axiom_status — stack status (version, tests, patents)
  • axiom_stop — stop-authority gate + real-time kill switch for irreversible actions

Enterprise

Everything in Core, plus:

Every tool result is HMAC-SHA256 signed under axiom-mcp-v1 — your editor can verify the signature client-side and detect any tampering between server and client. Additional runtime tools ship in the Axiom OS console; a signed-agent marketplace is on the way.

Full tool reference → Hosted manifest →

Pricing

Free

$0

  • 1,000 calls / month
  • Default policy
  • Community support

Team

$199/mo

  • 500,000 calls / month
  • Multi-tenant policies
  • Dashboard analytics

Enterprise

Custom

  • SOC 2, DPA, BAA
  • Dedicated tenant
  • SLA + support