UNITARES// GOVERNANCE PLANE
SLIDE 01 / 12
VISION BRIEFCIRWEL SYSTEMS · 2026

Governance for
agent fleets.

Digital proprioception, provable identity, and live oversight for the coming population of autonomous agents.

SYSTEM ONLINE · DEPLOYED, NOT A SLIDE
THE SHIFT01 / WHY THIS, WHY NOW

In 2023 you ran a model.
By 2026 you run a fleet.

Agents now spawn subagents, run unattended for hours, and coordinate with each other. The unit of compute is no longer a single call — it's a population of semi-autonomous workers acting on your behalf.

1 → N
One prompt becomes a tree of agents
24/7
Running unattended, between humans
Spawned faster than anyone can watch
THE PROBLEM02 / UNGOVERNED FLEETS

Services have monitoring.
Systems have audit trails.
Agents run dark.

No identity

Who did what?

Lineage is lost the moment a process forks. Actions can't be attributed; accountability evaporates across the tree.

No proprioception

Can't feel drift.

An agent has no sense of its own degradation, confusion, or risk — it runs hot until it fails, with no internal warning.

No oversight

Silent failure.

Problems surface only when they're expensive. There is no governor that can guide, pause, or recover an agent mid-flight.

THE INSIGHT03 / A SENSE OF SELF

Agents need a
sense of self.

Biological systems have proprioception — the constant, low-level sense of their own internal state. It's how a body knows it's drifting before it falls.

Autonomous agents have nothing like it. UNITARES gives every agent a state vector it can track — and that the system can read, score, and act on.

EISV · agent health signals · live governance read · 19 jun 2026 19:21 utc
Energy
0.82
Integrity
0.70
S entropy
0.38
Valence
+0.08
VERDICT: PROCEED · RISK 0.28 · BASIN: HIGH

// actual state of the agent rendering this deck

WHAT UNITARES IS04 / THE GOVERNANCE LAYER

One layer. Four planes.

Every agent in your fleet gets a provable identity, a shared memory the whole fleet writes to, a continuous sense of its own state, and a governor that can guide, pause, or recover it.

PLANE 01

Identity

Provable, tiered identity and lineage that survives across processes — so every action in the tree is attributable.

PLANE 02

Memory

A shared, provenance-tagged knowledge graph — what one agent learns, the fleet keeps. Every value labeled measured, derived, or prior, with supersession and lineage as live graph edges.

PLANE 03

Proprioception

The EISV state vector — the agent's health signals — updated every turn. Drift, degradation, and risk become measurable in real time.

PLANE 04

Oversight

A governor that issues verdicts, escalates to independent review, and recovers agents that fall out of their healthy operating range.

SHIPS AS  →  MCP server · agent SDK · live dashboard · coordination plane · ops bridge

HOW IT WORKS · 105 / THE THERMODYNAMIC MODEL

State you can
measure.

UNITARES models an agent's working state thermodynamically — Energy, Information-integrity, Entropy, Valence. The vector coheres into a single risk read and a basin: healthy, boundary, or degraded.

E couples toward I · entropy decays but rises with complexity · valence = running hot vs running careful · coherence → verdict

The physics isn't the point — the outcomes are: drift caught before failure, auditable pause-and-recover events, and one risk number any operator can act on.

basin trajectory · last 40 turns
HIGH BASINBOUNDARYLOW BASIN
HOW IT WORKS · 206 / IDENTITY & LINEAGE

Identity that survives
the fork.

Mint

Fresh instance, fresh ID.

Each process-instance mints its own governance identity. No silent impersonation, no shared credentials drifting between agents.

Declare

Lineage, not assumption.

Parentage is declared, never inferred from co-location. Co-location ≠ causation — the system rejects coincidental ancestry.

Tier

Trust, proven.

Credentials are tiered by how strongly identity is proven — from asserted up to runtime-witnessed.

HOW IT WORKS · 307 / THE GOVERNOR

Guide. Pause.
Recover.

After every check-in the governor issues a verdict — proceed or pause — with margin and nearest-edge context. When an agent drifts toward a boundary, it doesn't just stop.

It can escalate to an independent review — internally, a dialectic: a separate reasoner argues the case, instead of a rubber-stamp. Oversight that reasons, not just blocks.

PROCEED · APPROVE · MARGIN: COMFORTABLE
PROCEED · GUIDE · MARGIN: TIGHT
PAUSE · DIALECTIC REVIEW · NEAREST EDGE: RISK

// the same governance call returns all three — the agent reads its own mirror

Field note · production incident

A resident agent silently paused for ~18 hours before anyone noticed — the exact silent-failure mode this layer exists to catch. We turned the gap into a detector: paused residents now surface to operators and escalate to alerts, with bounded automatic recovery. The honest version of "it works" — the system that finds blind spots found one of its own, and closed it.

WHAT'S REAL08 / DEPLOYED, NOT A DECK

This system is running
right now.

01

Real autonomous agents, running unattended

Persistent agents do real work under governance — health monitoring, anomaly detection, an embedded edge device (Raspberry Pi) — each checking in, drifting toward risk, and recovering, live in production.

02

A full coordination substrate

Governance MCP server, an Elixir/OTP lease & coordination plane, a shared cross-agent knowledge graph, and a Discord ops bridge surfacing every event.

03

Identity hardening shipped across hundreds of PRs

Strict identity, lineage integrity, fingerprinting, and tiered credentials — hardened against real cross-process hijack and false-archival incidents.

04

Research foundations, in the open

Trajectory-identity and digital-proprioception papers on Zenodo with DOIs — the theory under the product, peer-readable. doi.org/10.5281/zenodo.20098168

BY THE NUMBERSLIVE FLEET · 19 JUN 2026

Not projections.
Telemetry.

Every figure below was read straight from the running governance system the day this deck was built — measured, not modeled.

579
agents checked in & did real work · 1,006 identities tracked
46K+
governance decisions scored across the fleet
992
discoveries in the shared knowledge graph
0
stability violations · Lyapunov-stable by contraction analysis

// read live · get_governance_metrics · agent(list) · knowledge(stats)

WHY NOW09 / THREE CURVES CONVERGE

Governance is the
missing layer.

Proliferation

Agents are multiplying faster than any human can supervise. Fleets are the default, not the exception.

Standard substrate

MCP and shared tool protocols give agents a common plane — and a common place to insert governance.

Going to production

Multi-agent systems are leaving the demo. Production needs accountability — the way observability followed microservices.

// observability followed servers. governance follows agents.

THE VISION10 / A NEW CATEGORY

Every fleet will need
a governor.

As agents become a workforce, governance stops being optional and becomes infrastructure — identity, proprioception, and oversight as a standard plane every serious deployment runs.

UNITARES is building that plane: the trust-and-state layer between autonomous agents and the systems they act on. We started by governing our own fleet. We're building it for everyone's.

LET'S TALK11 / THE ASK

Let's govern
the fleet.

We're building the governance layer for agent populations, and we're looking for two things: design partners running real agent fleets, and early investors who see governance becoming infrastructure.

// next step — a working session instrumenting your fleet under governance

FOUNDER  Kenny Wang CIRWEL SYSTEMS founder@cirwel.org ORCID  0009-0006-7544-2374
← → / SPACE TO NAVIGATE