DirigeX.AI   Verifiable AI Control Plane
Proven, not asserted →
The pivot to plug-and-play control

Own the layer that
outlives every vendor.

Dirigex decouples your governance, identity, policy, verification, and evidence layer from the agent vendors underneath it — so a vendor becoming a liability is an operational decision, not a governance rebuild.

$ npm i @dirigexai/sdk  ·  pip install dirigex-sdk

ConstantYours
◆ Control plane
Governance · Identity · Policy · Evidence
allow / require_approval / deny verified · expiring identity replayable evidence
Vendor AMCP
Vendor BA2A
Vendor Cliability
Vendor C′verified
Vendor DMCP
Replaceable · Volatile · What you route to
The wrong lock

You're worried about lock-in. You're pointing at the wrong lock.

The trap was never the agent you route to. It's the governance, identity, policy, verification, and evidence stack you'd have to rebuild the day you leave — slow-moving, expensive, legally load-bearing, and yours. Today it's welded to a vendor's control model: adopt their catalog, inherit their control model; leave the vendor, leave your control model behind. Lock-in isn't a pricing problem you negotiate out of. It's an architecture problem you design out of.

The honest split

Decouple your capabilities from your constraints.

Capabilities

The work — replaceable.

The agents and vendors that do the work. Commoditizing, volatile, and non-monotonic in quality. Replace them as the market moves.

What you route to →
Constraints

The control — durable.

Your governance decisions, verified identity, per-tenant policy, RBAC, scoped keys, and replayable evidence. Expensive to build, expensive to rebuild — and yours.

What you own ◆

The control plane sits above the vendor and stays constant no matter what runs underneath it. You build it once; it outlives every vendor you route through.

Stated plainly

What "swap" means here — and what it doesn't.

We say the boundary out loud, because for a trust brand, saying it is the trust.

What a swap means

  • Governance, identity, policy, verification, and evidence stay constant across a vendor change.
  • You sever a vendor at the radius you choose — global, tenant, or agent.
  • You register, verify, and route to an alternative — without rebuilding governance.
  • One evidence standard persists across the swap, so you can prove the new path meets the bar.

What it doesn't mean

  • No automatic failover, hot-swap, or self-healing routing. Substitution is a decision you make — never a silent system behavior.
  • No in-flight re-drive to a backup vendor. Retry is same-endpoint; the dispatcher's fallback is same-agent.
  • No "zero downtime." Routing is deterministic and manifest-pinned for reproducibility; when active state goes stale it fails closed and loud, it doesn't silently reroute.

The failover fantasy is easy to say and impossible to defend. Operational continuity of your control plane is hard to build, hard for a capability-layer competitor to claim, and provable — so that's exactly what we sell.

Buyer-grade, status-honest

The proof underneath the promise.

Governed capability recordPortable across vendors
01

Govern across pluggable catalog sources — not one vendor's.

One control surface, multiple protocols: MCP and A2A run end-to-end today; HTTP/JSON, gRPC, and WebSocket are supported at the platform layer — all on a pluggable registry source, where adding a catalog source is one implementation, not an ingestion rewrite.

02

Cut off a liability at three radii.

Global, tenant, or agent — sever at the radius you choose, and retire an agent's lifecycle outright. Kill switches built into the request path, not bolted on.

03

Identity travels with the record, not the vendor.

Agents carry verification tiers that expire; a freshness runner re-checks on a schedule and delists anything that fails. The trust doesn't reset when you change providers.

04

Every decision is replayable evidence — whoever you were routing to.

Each governed request persists a route trace with protocol, decision, and provenance, with configurable retention.

"If you can't prove how your AI decided, you can't defend what it did."

Where the lock-in actually lives

A catalog locks the capability layer. Dirigex keeps the control layer portable.

Registries make you pick a catalog.
Dirigex governs across pluggable catalog sources.
Registries give you a directory.
Dirigex gives you a decision — and the evidence of it.
Registries fuse capability and control.
Dirigex decouples them by design — what a capability-layer vendor structurally can't sell.

So the catalog becomes a choice you can change, not a commitment you can't.

The category move

Management is table stakes.
Portable control is the category.

Dirigex doesn't promise magic failover. It makes governance portable, evidence durable, and vendor severance an operational decision instead of a trust-stack rebuild.

✶ Proven, not asserted ✶