Skip to content

Vargate Developer Docs

Vargate builds independent, cryptographically verifiable governance for AI. We run two products under one thesis: the customer should own the record.

  • Tyr governs autonomous agents inline. It sits between your agent and the tools it calls, evaluates every tool call against policy, and blocks, allows, or escalates before the action happens.
  • Ogma audits human and API AI usage independently, across vendors — Claude and OpenAI today. It pulls from each vendor's management APIs with an admin key — no agents, no proxies — and hash-chains every event into a tamper-evident record that lives outside the vendor's perimeter.

Both products write a hash-chained, blockchain-anchored audit trail. Tyr prevents bad actions; Ogma proves what happened.

Choose your product

  • Tyr — govern agents inline

    A supervision proxy for autonomous agents. Policy enforcement, human-in-the-loop approval, and a verifiable audit trail for every tool call.

    Tyr quick start

  • Ogma — audit usage across vendors

    The independent audit layer for human + API AI usage. Connect Claude and OpenAI with an admin key; every event is chained and anchored.

    Ogma overview

  • Get started in under 10 minutes

    Sign up, get your API key, and send your first governed action.

    Quick Start

  • API Reference

    Full endpoint documentation with request/response schemas.

    API Overview

  • Install the CLI

    pip install vargate-cli — manage your governance proxy from the terminal.

    CLI Guide

  • Policy Templates

    Pre-built governance policies for financial, email, CRM, data access, and general use cases.

    Browse Templates

  • Managed Agents

    Govern Anthropic managed agents with active policy enforcement, passive observability, and session lifecycle control.

    Managed Agents Guide

How the two products differ

Tyr Ogma
What it governs Autonomous agents you run Human + API AI usage across your org
Where it sits Inline — between the agent and its tools Outside the vendor — reads management APIs
Posture Active: blocks, allows, or escalates before an action runs Passive: observes, attributes, and flags after the fact
Integration Your agent calls the Tyr gateway Paste a vendor admin key; no infrastructure changes
Vendors Any agent that routes tool calls through the proxy Claude (Anthropic) and OpenAI today
Shared foundation Hash-chained, blockchain-anchored audit trail Hash-chained, blockchain-anchored audit trail

Most regulated teams run both: Tyr keeps autonomous agents inside their lane, and Ogma builds the independent ledger of everything your people do with AI.

Key capabilities (Tyr)

Policy-Based Governance (OPA/Rego)

Every tool call is evaluated against OPA/Rego policy before execution. Two-pass evaluation: a fast path for simple decisions, and an enriched path that incorporates behavioral history and anomaly scoring.

Hash-Chained Audit Trail

Every decision is written to a hash-chained audit log. Each record's hash includes the previous record's hash, making the trail tamper-evident. Verify integrity at any time with a single API call.

Merkle Tree Aggregation + Blockchain Anchoring

Audit records are aggregated into hourly Merkle trees with O(log n) inclusion proofs. Tree roots are anchored to Polygon and Ethereum, creating an independently verifiable, immutable record.

Human-in-the-Loop Approval

When policy requires human review, the action is queued — not executed. A human approves or rejects, and only then does execution proceed. Full audit trail of who approved what and when.

Crypto-Shredding (GDPR Erasure)

PII in action parameters is automatically detected and encrypted with per-subject HSM keys. GDPR erasure destroys the key, rendering all ciphertext for that subject irrecoverable — without breaking the hash chain.

Webhook Notifications

Get notified instantly when actions are denied, escalated, or approved. HMAC-SHA256 signed payloads with retry and exponential backoff.