Pattern matrix/White paper/Governance

ADPS Agent Design Pattern White Paper · Module overview

Governance Module: Making Agent Autonomy Manageable

Authorization, accountability, containment, cross-cutting controls, lifecycle, and control plane.

When an agent only recommends, governance often looks like content review. Once it invokes tools, changes business state, and delegates to other agents, governance becomes a runtime discipline: whom the agent represents, why an action is authorized, where failure stops, and whether responsibility and outcome can be reconstructed.

Locally compliant steps can still drift away from a long-running goal. Governance therefore evaluates both the current action and its relation to the original goal and real outcome.

ADPS v0.4 dual-axis matrix, cross-cutting engineering planes, and agent lifecycle

Three governance objectives

ObjectiveQuestionEngineering objects
AuthorizationWho represents whom, and which action may affect which resource?Identity, delegation, policy, tool, arguments, resource
AccountabilityWho did what under which versions and policies, and what happened?Run, trace, approval, versions, state delta, external receipt
ContainmentIf a control fails, where does the maximum impact stop?Sandbox, tenant boundary, quota, budget, breaker, compensation

G1 admits one intent. G2 bounds the damage if an admitted action is wrong. G3 changes the authority of a capability over time. G4 supplies their evidence, and G5 provides deterministic enforcement points.

The v0.4 structure

Dual-axis matrix

Cognitive function × execution topology remains the primary ADPS structure. Seven functions and six topologies are unchanged. The matrix now contains 27 cell-bound core patterns; the Governance row retains G1 Approval Gate, G2 Blast-Radius Control, and G3 Progressive Commitment.

Cross-cutting engineering planes

PlaneScopeOutputs
Observation and evidenceEvery pattern and lifecycle stageEvents, causality, versions, state deltas, receipts
Evals and testingArtifacts, trajectories, business outcomesRegression, graders, deterministic tests, acceptance
Identity, policy, and securityPrincipal, delegation, authority, boundariesAllow, deny, ask, limits, execution conditions
Human-agent interactionClarification, approval, pause, takeover, resumeTyped events, review context, recovery coordinates

G4 moves out of Governance × Orchestrate and becomes the twenty-eighth core specification: Cross-cutting core · Observation and evidence plane. The total remains 28: 27 matrix patterns plus one cross-cutting core.

Lifecycle

Agent governance lifecycle from registration and evaluation to demotion and retirement

ReAct is the perception-reasoning-action micro-loop inside controlled operation. Evals recur during offline validation, canary release, operation, and revalidation. Neither requires another matrix coordinate.

Running example: a payroll batch across the lifecycle

This example explains the composition and is not an attributed enterprise case.

StageEngineering actionEvidence
Register and ownRecord the agent, owner, purpose, production environment, two capabilities, and retirement conditionsAgent ID, owner, capability list, credential references
Design and versionPin model, prompt, payroll rules, payment tool, approval policy, and data dependenciesWorkload digest and dependency manifest
Offline evalCover routine batches, new hires, cross-region tax, duplicate requests, and hostile argumentsRegression results, failure classes, cost, boundary tests
Shadow and canaryCompare with the manual process before enabling a small recommendation sliceHuman differences, adoption, long-tail distribution
Controlled operationG1 freezes and routes intent; G2 limits tenant, amount, and batch; G5 revalidates before commitIntent, Approval, quota, hook verdicts
Observe and attributeG4 links source ledger, arguments, payment receipt, state delta, and reconciliationCross-system trace, receipt, business outcome
Change and revalidateA payment-API upgrade freezes old grants and reruns affected evaluation and shadow workVersion diff, revalidation, change approval
Authority dispositionValidation may reach bounded execution; submission remains reviewed; an ended pilot is retiredCapability grant, demotion, or retirement record

Governance control plane

Agent governance control plane with registry, policy, enforcement, and observation

A local middleware can guard one agent. Multiple agents, domains, and delegations need shared registry, policy, enforcement, and evidence services. A central platform can own identity, policy formats, telemetry, and cross-domain audit; domain teams still own risk, acceptance, approvers, and incident response.

Governance contract

intent_id: int_01K3...
principal: user://finance/108
agent: agent://payroll/prod-v7
run_id: run_8842
tool:
  name: create_payment_batch
  digest: sha256:4ef...
resource_scope:
  tenant: tenant_42
  max_records: 20
policy:
  version: payroll-policy-v12
  decision: ask
preconditions:
  source_ledger_version: 417
approval:
  expires_at: 2026-08-19T09:30:00Z
  max_uses: 1
execution:
  idempotency_key: payrun-2026-08-batch-17

A tool name is not enough for a governance decision. Immutable tool version, canonical arguments, resources, delegation, environment, aggregate budget, and business preconditions all affect risk.

Pattern responsibilities

SpecificationControl objectBoundary
G1 Approval GateCurrent high-risk intentRevalidate intent and preconditions on resume; consume approval once
G2 Blast-Radius ControlMaximum action, run, and fleet impactAgent judgement and historical success cannot alter hard limits
G3 Progressive CommitmentCapability- and scenario-specific autonomyPromote, hold, demote, freeze, and retire
G4 ObservabilityCross-cutting evidence chainKeep observed facts separate from evaluation judgements
G5 Hooks PipelineDeterministic enforcement pointsA hook is not the policy source or domain judge

Public engineering material

DeerFlow Guardrail and two-layer authorization follows five public pull requests through pre-call interception, principal propagation, RunJournal, an RBAC provider, and assembly-time plus runtime authorization.

First Governance Module Workshop preserves concrete questions about sandboxing, approval resume, agent registry, evidence attribution, capability-level authority, and offline evaluation, then records their effect on v0.4.

Mechanisms under further study

  • Delegated identity chain across user, agent, workload, run, and downstream tool.
  • Durable intent and the business changes that invalidate an approval during a wait.
  • Agent registration and retirement for pilots, credentials, versions, queues, and ownership.
  • Federated governance between central controls and domain accountability.

Suggested citation:ADPS, Governance Module: Making Agent Autonomy Manageable, Agent Design Pattern White Paper v0.4, 19 August 2026.

Pattern catalog · Governance workshop · CC BY 4.0

Scope:Public review draft. Definitions and classifications are open for discussion and citation; running examples explain mechanisms, while attributed practice appears in the case library.