Agiorcx

One console for every agent your business runs.

The Agiorcx platform sits above the SDK and gives your team full visibility and control — policy enforcement, HITL approval flows, coordination contracts, and an immutable audit ledger.

Agent Runtime

Execute agents reliably at scale. The runtime enforces policy sets, manages concurrency, handles retries, and emits structured audit events for every action taken.

Policy enforcement at execution time — not configuration time
Concurrent execution with built-in conflict detection
Automatic retry with backoff and full failure attribution
Structured event emission for every agent action

Agent Identity

Every agent gets a unique, verifiable identity. Actions are attributed to the exact agent instance, version, and policy set it ran under — not just the model.

Unique identity bound to agent instance and version
Identity scoped to policy set and workspace
Attribution on every action in the audit ledger
Identity verification for cross-org communications

Governance Engine

Define, version, and enforce policies across every agent in your workspace. Set rate limits, scope boundaries, and allowed action sets. Violations are logged and surfaced in real time.

Declarative policy definitions — version-controlled
Rate limits, scope boundaries, and action allowlists
Policy inheritance across workspace and project levels
Real-time violation detection and alerting

HITL Inbox

Every agent action requiring approval surfaces in a structured inbox. Review, approve, or reject with full context — before anything executes. Human-in-the-loop as a first-class primitive, not a plugin.

Approval inbox with full action context and provenance
Configurable approval thresholds per action type
Time-boxed approvals with automatic escalation
Bulk review for repeated action patterns

Audit Log

An immutable, append-only ledger captures the full action trail. Every step attributed, every decision logged — with enough context to replay, debug, or export for compliance.

Immutable, append-only event ledger
Structured event schema with full provenance
Replay and diff tooling for debugging
Compliance export in standard formats

Connector Interface

Cross-organisation trust agreements, federation channels, and agent-to-agent communication. Agents from different organisations negotiate scope and execute within governed contracts.

Cross-org trust agreements and federation channels
Contract negotiation protocol for agent-to-agent scope
Governed communication with attribution on both sides
Webhook and event delivery for external integrations

SDK at the edge. Platform at the centre.

The SDK runs in your infrastructure and handles agent execution. The platform handles everything above that — governance, approval, coordination, and audit.

You own the agent runtime. We own the control plane. Your data never leaves your infrastructure unless you explicitly route it through the platform.

Console UIBrowser-based management and approval interface
Platform APIGovernance, HITL, and coordination endpoints
Policy EnginePolicy evaluation, enforcement, and violation logging
Coordination BusContract negotiation, session management, state
Audit LedgerImmutable, append-only event log
@agiorcx/sdkOpen source agent runtime — runs in your infrastructure

↑ Platform (hosted)    ↓ SDK (your infrastructure)

Ready to govern your agents?

Platform access is invite-only and free during alpha.

Request access