Run Aegisora on your own agents in a scoped, time-boxed pilot.Apply now
Technical documentationRuntime v4.0.0

Aegisora runtime reference

How Aegisora governs autonomous AI agents: the decision pipeline, canonical authority, execution continuity, evidence and enterprise controls.

12
pipeline stages
3
decisions: ALLOW, BLOCK, ESCALATE
9
aligned packages (v2.0)
115/115
runtime tests (v4.0.0 notes)
Introduction

Overview

Aegisora is a runtime governance layer that evaluates and enforces agent actions before they cross protected execution boundaries.

Securing the model is not enough once agents can call tools, providers and external systems. Aegisora secures what the agent is allowed to do: every action is evaluated, decided and enforced at the execution boundary, and every outcome is recorded as evidence.

Enforcement
Governed before executed

A blocked or escalated action never reaches the provider or the tool.

Zero trust
Fail closed

When state, identity or authority no longer matches what was authorized, execution stops.

Audit
Evidence by default

Allowed, blocked and escalated actions all produce correlated, auditable records.

Enterprise
Tenant isolated

Workspace isolation, RBAC and server-side authorization are part of the runtime.

Where Aegisora sits

LayerRole
Aegisora RuntimeSecurity, governance, policy, risk, decisions, enforcement.
Agent frameworkOrchestration and workflow execution (for example LangChain and LangGraph).
Model providerModel inference and generation.
Tools and APIsExternal execution surfaces governed by Aegisora.
Designed to fit the existing stack

Aegisora works with the AI stack developers already use rather than replacing it. It sits between the agent and the execution surfaces it can reach.

Introduction

Governance pipeline

Every governed action follows one canonical chain, from request to audit. Select a stage to see what it guarantees.

Stage 7 of 12Introduced in v4.0

CONTINUITY SEAL

The authorized action is bound to the security state it was approved under.

  • Binds workspace, agent, execution identity, action, resource, tool, execution input and authority state.
  • The seal has an immutable identity and cryptographic integrity validation; inputs are canonically normalized.
  • A seal cannot be created or executed while a containment condition is active.
New in v4.0

Continuity Seal and Effect Reconciliation extend the chain from deciding whether an action is allowed to proving the executed action is still the authorized one.

Core concepts

Decisions and enforcement

The security boundary is explicit: ALLOW, BLOCK or ESCALATE.

DecisionEffect at the execution boundary
ALLOWExecution proceeds and carries canonical correlation into the governed execution.
BLOCKProvider and tool execution is prevented. The prevented, not-attempted state is preserved.
ESCALATEExecution stops pending approval. Provider and model resolution do not occur.

Enforcement state and execution outcome

Aegisora separates what governance decided from what actually happened. This lets you answer both questions for any action.

DimensionValues
Enforcementnot_executed, executed, prevented, escalated
Executionnot_attempted, succeeded, failed

Canonical correlation identifiers

IdentifierPurpose
traceIdTies one request across governance, enforcement, execution, evidence and audit.
decisionIdIdentifies each governance decision independently.
executionIdIdentifies the governed execution.
evidenceIdIdentifies the evidence record.
correlationIdRuntime-generated correlation across the security record.
sealIdIdentifies the Continuity Seal (v4.0) and links it to evidence.
Identity is runtime-generated

Attacker-controlled correlation identifiers are rejected as canonical identity. Forged runtime identities fail closed.

Core concepts

Canonical authority

One authority speaks for governed execution, and execution requires a valid, scoped capability.

ToolRegistry is the single canonical authority for governed tool execution. Overlapping legacy authority surfaces were removed so that execution can never be governed by competing implementations.

Authority chain
GOVERNANCE
  |
  v
ToolRegistry
  |
  v
Authorization Capability
  |
  v
AgentExecutor
  |
  v
RuntimeTool
  |
  v
EXECUTION

Capability-based execution

  • Authorization must exist and be valid.
  • Authorization must match the scope of the requested action.
  • Authorization is single use, which provides replay resistance at the execution boundary.

Provider governance ordering

Resolution order
REQUEST -> GOVERNANCE -> PROVIDER RESOLUTION -> MODEL RESOLUTION -> PROVIDER INVOCATION

Governance runs before provider and model resolution. BLOCK and ESCALATE therefore prevent resolution entirely. Provider identity and model identity are immutable after governance, and provider and model must be consistent.

Runtime security

Continuity Seal

Authorization answers whether an action is allowed. Continuity answers whether it is still the same action, under the same authorized security state, at the moment of execution.

Introduced in v4.0.0

Continuity Seal extends runtime governance from authorization and policy enforcement into the integrity of the execution itself.

What a seal binds

BindingVerified at execution time
Workspace identityThe execution remains in the authorized workspace.
Agent identityThe same agent that was authorized is executing.
Execution identityThe execution matches the sealed execution identity.
Action, resource, toolThe action, target resource and tool have not changed.
Execution inputExact input is verified after canonical normalization.
Authority stateEffective authority has not mutated or drifted since authorization.
Containment stateNo containment condition is active.
Seal integrityThe seal itself passes cryptographic integrity validation.

Authority integrity

  • Authority-state fingerprinting
  • Authority mutation detection
  • Authority drift detection
  • Transitive authority reconciliation
  • Workspace-scoped authority verification
  • Fail-closed execution when authority changes

Enforcement at the boundary

Continuity at the execution boundary
VALID STATE            CHANGED STATE
    |                      |
    v                      v
  ALLOW                  BLOCK
    |                      |
    v                      v
EXECUTION          NO EXTERNAL SIDE EFFECT

A continuity violation is enforced before the external side effect occurs. Modified action payloads, resources, tools or execution inputs are rejected when they no longer match the sealed execution state.

Provider and tool execution

Provider security
Providers

Provider execution sealing, identity binding, runtime verification before invocation, input integrity and fail-closed enforcement.

Tool security
Tools

ToolRegistry authorization binding, continuity-aware receipts, execution-time seal verification, exact input verification and authority mutation blocking.

Effect reconciliation

After execution, the runtime reconciles the observed effect against the sealed execution identity: authorized intent, sealed execution, actual effect. The reconciliation result is part of the evidence record.

Runtime security

Evidence and audit

Evidence is a first-class security object. Every decision leaves a record, including the ones that never executed.

  • Canonical evidence creation at the enforcement boundary
  • Persistent evidence storage
  • Deterministic trace, decision, execution and evidence correlation
  • Synchronized evidence finalization
  • Retention for blocked and prevented actions
  • Retrieval by trace, decision and agent
  • Execution-level linkage through sealId and continuity enforcement status (v4.0)
Blocked actions are evidence too

A BLOCK is not silence. The prevented action is retained with its decision, so an auditor can see what was attempted and why it was stopped.

The enterprise layer adds an immutable evidence ledger and an enterprise audit ledger, with trace, decision, execution and evidence correlation preserved end to end.

Enterprise governance

Enterprise governance

The enterprise runtime introduced in v3.0 adds identity, access, registries and policy management around the core pipeline.

AreaCapabilities
Identity and accessWorkspace isolation, RBAC, server-side authorization, tenant boundary enforcement
Agent and provider governanceAgent registry and discovery, provider and model registry, governed model resolution, provider and model identity integrity
Policy and riskPolicy Studio, policy versioning, risk detection, deterministic decision precedence
Tool securityTool security evaluation, ToolRegistry enforcement, external tool governance, MCP tool governance
Zero-trust execution

The runtime is designed around zero-trust execution, tenant isolation, auditable decisions and evidence-backed enforcement.

Enterprise governance

Approval engine

ESCALATE decisions stop execution until a person or process resolves them.

Approval lifecycle
PENDING -> APPROVED
        -> REJECTED
        -> EXPIRED
  • Approvals are workspace-isolated.
  • Approval state is enforced at runtime, not only recorded.
  • Resolution is synchronized back to the governed execution.
Enterprise governance

Tool and MCP security

Tools are the point where an agent's intent becomes a side effect, so they are governed by the same pipeline as providers.

  • Tool security evaluation before authorization
  • ToolRegistry enforcement as the single canonical authority
  • External tool governance
  • MCP tool governance
  • Authorization receipts bound to the Continuity Seal (v4.0)
  • Post-execution effect reconciliation (v4.0)
Enterprise governance

Operations and entitlements

Operational surfaces that let a security or platform team run governance day to day.

Realtime events

Enterprise realtime events, a runtime-to-enterprise event bridge, persistent realtime writers and event subscriptions.

Incidents and alerts

Enterprise incident ledger, governance orchestration, alert rules and webhook governance.

Usage and entitlements

Usage ledger and metering, persistent usage projection, execution entitlements, provider-level entitlement and execution quota enforcement.

Contracts
Enterprise contracts

Billing state, SSO identity, SCIM provisioning and enterprise recovery contracts.

Platform

Packages

As of v2.0.0, nine packages are aligned to a single version.

PackageArea
@aegisora/coreCore types and shared foundations
@aegisora/runtimeGoverned execution runtime
@aegisora/policy-enginePolicy evaluation
@aegisora/security-engineSecurity and risk analysis
@aegisora/auditAudit records
@aegisora/storagePersistence
@aegisora/observabilityRuntime observability
@aegisora/pluginsExtension points
@aegisora/sdkDeveloper SDK
Area descriptions

Areas are summarized from package names and release notes. Consult each package README in the core repository for exact APIs.

Platform

Integrations

Aegisora intercepts agent actions rather than replacing the framework that produces them.

IntegrationStatus
LangChain and LangGraphSupported since v0.1.0 for runtime request interception
MCP toolsGoverned through tool security evaluation (v3.0)
Model providersGoverned provider and model resolution; Anthropic is documented on docs.aegisora.org
External tools and APIsGoverned through ToolRegistry authorization
Platform

Getting started

The smallest complete governance flow, and the rules for configuring it safely.

Smallest governance flow
Agent -> Request -> Governance Runtime -> Decision -> Execution

Configuration

Configuration covers runtime behavior, policies, providers, logging and environment variables. Never commit API keys or other credentials; use local environment configuration for development secrets.

Runnable examples

Maintained executable examples live in the examples directory of the core repository and run as a smoke test on every release.

Platform

Release validation

Release claims are tied to checks that run against the actual artifacts.

v4.0.0 validation, as published in the release notes

CheckResult
Continuity Seal adversarial tests20/20 passed
Runtime continuity boundary tests5/5 passed
Runtime tests115/115 passed
Build, typecheck, runtime typecheckPassed
Example smoke testPassed
Security Readiness, CodeQLPassed
Post-merge main validationPassed
Provider boundary and ToolRegistry receipt validationPassed
Authority and input mutation blocking, effect reconciliationValidated

Supply-chain discipline

  • Security Readiness enforced in CI
  • Package metadata, npm tarball content and package scope validation
  • Release artifacts regenerated from the final release tree, with SHA256 evidence (v2.0.0)
  • Release gate metadata rebound to the final commit
Platform

Release history

From a governed-execution foundation to execution continuity.

VersionDateNameFocusCommit
v4.0.0Sep 2026Continuity SealExecution continuity and runtime securitya99fa63
v3.0.0Sep 10Enterprise Governance RuntimeWorkspaces, RBAC, approvals, usage, entitlements338fbce
v2.0.0Aug 30Governance PlatformNine aligned packages, unified platform08033e3
v1.8.0Aug 30Runtime HardeningCorrelation identity integrity99da016
v1.7.0Aug 30Governance BoundaryBoundary contract, fail-closed identities94521df
v1.6.0Aug 30Canonical AuthorityToolRegistry as single authority3090fd5
v1.5.0Aug 29Evidence IntegrityEvidence as a first-class security objecte24aee3
v0.1.4Aug 26Release HardeningSecurity Readiness CI, package validation02e5479
v0.1.2Aug 21StabilizationPackage organization, developer onboarding01a74b7
v0.1.0Aug 18FoundationFirst governed-execution patha494e78

Each release is tagged in the core repository. Compare views and full notes are available on the GitHub releases page.

This reference is derived from the published Aegisora release notes through v4.0.0 and the public documentation at docs.aegisora.org. Exact APIs and runnable code live in the core repository.

Security policy ยท Contributing