Introduction
AuthSpoke is the Enterprise AI Control Plane: a headless, API-first platform for AI assets, agents, connectors, governance, policy, risk, discovery, and audit.
AuthSpoke is the Enterprise AI Control Plane. It gives every AI agent, MCP server, model, tool, connector, and imported provider asset a governed place in the enterprise: identity, ownership, lifecycle, metadata, policy, review, diagnostics, synchronization, and audit.
It is not another IAM product. Okta, Microsoft Entra, and SailPoint govern people and workforce access. AuthSpoke governs autonomous AI systems and the enterprise connectors that discover, import, provision, synchronize, and evaluate them.
API-first by design#
The backend is the product. The console, MCP servers, AI agents, server-to-server jobs, CI workflows, and future SDKs all use the same documented APIs:
- No tenant in the API path. Tenant context is resolved from the bearer token/session.
- Headless endpoints. Use the same
/api/v1/...resources from UI, automation, AI agents, MCP servers, or machine clients. - Server-side governance. Risk, trust, lifecycle, policy findings, permissions, and audit records are computed and enforced by the backend.
- OpenAPI reference. The API Reference is generated from the current AuthGateway controller surface and includes the full API inventory.
What you can build#
- Register and govern AI agents, MCP servers, models, tools, identities, sessions, and compliance controls.
- Import provider-discovered AI assets into the AI Registry through the connector framework.
- Manage governed AI assets: ownership, lifecycle, governance state, metadata, relationships, activity, and cleanup.
- Define AI Governance policies, run evaluations, review findings, create exemptions, and track remediation tasks.
- Browse connector manifests, create connector instances, validate configuration, run diagnostics, import assets, synchronize drift, inspect operations, and query activity.
- Operate support/provider workflows: tenant lookup, entitlements, kill switches, runtime config, and controlled support actions.
API domains#
| Domain | Primary routes |
|---|---|
| Authentication and registration | /api/v1/auth/*, /api/v1/register, /api/v1/email/* |
| Enterprise AI | /api/v1/ai/*, /api/v1/ai-assets/*, /api/v1/ai-governance/* |
| Enterprise connectors | /api/v1/connectors/*, /api/v1/connector-instances/*, /api/v1/integrations/* |
| Platform and tenant | /api/v1/platform/*, /api/v1/tenant/domain/* |
| Provider support | /api/v1/support/* |
| Runtime SSO compatibility | /app/*, /callback |
Request model#
For tenant-scoped calls, send:
curl "https://api.authspoke.com/api/v1/ai-assets" \
-H "Authorization: Bearer $TOKEN"
The same call can also go through a tenant console host:
curl "https://your-company.authspoke.com/api/v1/ai-assets" \
-H "Authorization: Bearer $TOKEN"
Both forms call the same backend resources. The token/session decides the tenant and permission envelope.
Next steps#
- API Reference - the full AuthGateway API surface.
- Build an AI Agent - register, enrich, govern, evaluate, and audit an agent.
- Enterprise AI Control Plane - how the system fits together.
- Meridian scenario - a worked enterprise adoption path.