Features

Built for production agent security.

Cryptographic identities. Granular permissions. Immutable audit trails. Every feature designed with security, performance, and compliance in mind.

Ed25519Key Standard

Cryptographic Agent Identity

Every agent gets a unique Ed25519 keypair on registration. The private key is returned once and never stored. Public keys and SHA-256 fingerprints enable trustless verification across your entire infrastructure.

Ed25519 key generation

SHA-256 fingerprinting

One-time private key delivery

PEM format export

<5msVerification

Fine-Grained Permission System

Control exactly what each agent can do with 7 hierarchical scopes. Set time-based validity windows, rate limits, and conditional constraints. Permissions are cached in Redis for sub-5ms verification.

7 OAuth-style scopes

Time-based validity windows

Rate limiting per scope

Redis-cached checks

SHA-256Content Hashes

Immutable Audit Trail

Every agent action is recorded with ULID-ordered events, SHA-256 content hashes, and optional cryptographic signatures. TimescaleDB provides efficient time-series storage and querying.

ULID event ordering

SHA-256 content hashes

Batch event ingestion

Time-series optimized storage

allow / deny / escalatePre-action decisions

Runtime Control

Check every sensitive action before it runs. One SDK call returns allow, deny, escalate, or a modified payload against your live policy. High-risk actions route to a human approval queue, and every decision is recorded for audit.

Pre-action policy checks

Human approval queue for escalations

Modified-payload responses

Fail-open or fail-closed per org

Platform Capabilities

And Much More

Every detail considered for production deployment.

Trust Scoring

Multi-factor algorithm scoring agents 0-100 based on status, age, success rate, and scope breadth.

Real-Time Streaming

WebSocket-based audit event streaming with channel subscriptions for live monitoring.

Public Trust API

Rate-limited public endpoints for third-party trust verification and embeddable SVG badges.

API Key Management

Scoped API keys with SHA-256 hashing. Create, list, and revoke keys through the dashboard.

Multi-Tenant Isolation

Organization-scoped data with row-level security. Every query is automatically tenant-isolated.

TimescaleDB Integration

Automatic hypertable creation for audit events. Efficient time-range queries and data retention.

TypeScript SDK

Full-featured @vorim/sdk with typed methods for agent registration, permission checks, audit submission, and trust verification.

Python SDK

Synchronous and async clients via the vorim PyPI package. Full feature parity with the TypeScript SDK, including framework integrations.

Framework Integrations

Built-in support for LangChain, OpenAI, Anthropic/Claude, CrewAI, and LlamaIndex. Automatic permission checks and audit trails on every tool call.

MCP Server

Model Context Protocol server for Claude Desktop, Cursor, and VS Code. 19 tools — agents can register, check permissions, and log actions via natural language.

Cloud-Native Infrastructure

Containerised microservices with PostgreSQL, Redis, Kafka, and a real-time dashboard — fully managed so you can focus on building.

Compliance Exports

Generate signed audit bundles in JSON, CSV, or PDF with SHA-256 manifests for regulatory reporting.

Get Started

Ready to Secure Your Agents?

Start with the free tier. Add identity, permissions, and audit trails to your AI agents in under five minutes.