0% read on this device
Browse the curriculum

Start Here

Vision

First Principles

Operating Model

Domain Model

Agent Factory

Runtime Architecture

AI Engineering

Autonomous Workflows

Verification & Delivery

Factory Platform

Quality Engineering

Security & Governance

Case Studies

Labs

Interview Practice

Research Journal

Reference

Curriculum/Runtime Architecture/A focused view of boundaries, contracts, state, authority, failure paths, and tradeoffs drawn from this chapter.
Runtime Architecture10 min readchapterQuick Read

AI Software Factory Reference Architecture

Connect intent, authority, execution, verification, delivery, operations, and learning through one canonical architecture.

Status: Review readyRisk: highLifecycle: executeContent reviewed 2026-08-30Maturity guide →
Claim boundaryThis chapter references implementation evidence. Inspect its evidence boundary before treating a claim as proven.
architecture mode

A focused view of boundaries, contracts, state, authority, failure paths, and tradeoffs drawn from this chapter.

Whiteboard exercise

Reconstruct and defend this chapter’s architecture.

Reconstruct the architecture, name each boundary, and defend the tradeoffs.

runtimeharnessenvironment10 min chapter
Open the source exercise

Draw all planes and the golden path in ten minutes. For every arrow state the command/event, identity, authoritative record, failure mode, idempotency strategy, and human decision. Then redraw it as a three-deployment V1 and defend what you combined.

3.3 Deployment and trust-boundary view

The diagram is a scale-out option, not a service mandate. A V1 may combine API, orchestrator, policy, and records in a modular deployment. It must still use authenticated interfaces, separate worker and validator contexts, scoped credentials, explicit egress, immutable subject versions, and durable state.

Trust changes at channel-to-API, control-to-worker, sandbox-to-gateway, retrieval-to-context, producer-to-validator, evidence-to-decision, and delivery-to-production boundaries. Each crossing authenticates both sides, authorizes the exact action, validates schema and classification, limits tenancy and destination, and records a correlation key.

3.4 Authority and evidence view

RecordProduced byAuthorizes or provesMust bind
Intent decisionNamed business authorityWhy work may beginPurpose, outcome, constraints, owner
Policy decisionPolicy service under delegated governanceWhether a requested action is eligibleActor, subject, action, resource, context, policy version
GrantCredential or grant serviceNarrow runtime authorityRecipient identity, scope, purpose, expiry, revocation
Work order and manifestControl planeFrozen execution requestPlan, capabilities, context policy, budgets, quality contract
AttemptOrchestrator and workerWhat execution occurredManifest, worker, environment, lease, tool calls, artifacts
Evaluator resultIndependent quality contextOne measured claimSubject digest, evaluator version, dataset, result, uncertainty
Proof packageEvidence serviceEligibility for a named decisionAll required evidence, contradictions, expiry, lineage
Human decisionNamed decision ownerApproval, rejection, restriction, exception, or acceptanceExact subject, evidence, reason, conditions, expiry
Release recordDelivery systemWhat entered an environmentArtifact, configuration, migration, target, rollout, rollback
Outcome assessmentOperations and business ownerWhether value and safety held in realityRelease, observation window, SLOs, business outcome, incidents
Learning decisionChange ownerWhether a configuration change may be promotedBaseline, candidate, evaluation, risk, approval, rollback

Decision lineage is a graph of observable records. It intentionally excludes private chain-of-thought. Reproducibility comes from exact inputs, versions, actions, outputs, policy decisions, and evidence.

3.5 Boundary contract

Every material arrow in any view must define:

FieldRequirement
DirectionCommand requests a state change; event reports an accepted fact
IdentityAuthenticated caller, workload, recipient, and delegated authority
SchemaVersioned payload with classification and tenant
AuthorityOwning policy decision and permitted subject/action/resource
StateExpected version, preconditions, valid next states, authoritative store
DeliveryIdempotency key, deduplication, acknowledgement, deadline
FailureTimeout, retry class, backoff, circuit break, compensation, escalation
EvidenceCorrelation, input/output digest, audit event, proof eligibility
Human decisionNamed gate, options, required evidence, expiry and exception path
CompatibilityProducer/consumer support window, deprecation, revocation

Retries are safe only when the operation is idempotent or a reconciliation protocol can determine the external result. A timeout creates uncertainty; it does not prove failure. Critical boundaries fail closed when authority cannot be established and degrade explicitly when a safe read-only path exists.

3.6 Failure trace

For a compromised capability: registry revocation blocks new resolutions; the control plane identifies affected manifests and pauses work; grant and tool gateways revoke authority; workers checkpoint and stop; evidence derived from the version becomes ineligible; delivery blocks affected artifacts; incident response preserves state; recovered work resolves a qualified replacement, re-executes required checks, and requires a new decision. No single log or dashboard substitutes for this state reconciliation.

4. Tradeoffs

Plane separation improves reasoning and security but adds contracts and operational overhead. A modular monolith is often the right V1: one deployment, explicit modules, separate identities for external execution, and stable events. Microservices should follow demonstrated scale or isolation needs, not the diagram.

Multi-agent orchestration is a capability, not a mandatory topology. A single executor is preferable when specialization adds coordination cost without independent assurance.

5. Current Mission Control Implementation

Mission Control’s React UI and Convex functions form much of the control/data plane. The Hono orchestration server, executor adapters, worktrees, GitHub App integration, leases, run events, and artifacts form an emerging execution plane. QC runs, verification receipts, approval decisions, GitHub checks, and release-gate automation form an incomplete quality/delivery plane. Identity, permissions, governance policy, and sandbox controls supply parts of the security plane.

The boundaries are not equally mature. convex/qcRuns.ts still invokes mock adapters and contains a policy TODO; convex/governance/releaseGateAutomation.ts operates in SHADOW; PR publication and remote sandbox controls live on study branches rather than the cited main baseline; and the browser-operated Mission-to-validated-PR path remains the decisive proof gap.

Evidence boundary

Curriculum maturity is not implementation proof.

This chapter defines architecture or practice. It does not by itself prove a corresponding production implementation.

CurriculumReview readyImplementation evidenceNot asserted hereInspect evidence map →
External review

Review this chapter.

Challenge a claim, boundary, missing failure mode, unclear term, or unsupported evidence statement.

  • Claim
  • Boundary
  • Failure
  • Evidence