The vocabulary · one term, one definition, one anchor

Canonical vocabulary of the Agentive World

The terms of the trilogy, one line each, exactly as defined by the book that coins them.

Editions: I v1.0 · II v1.1 · III v1.2

This page defines nothing: each entry is read from the glossary of its book and links back to it. That is why it can be cited with confidence — the definition you see here is the one in the work, and each term’s anchor (#nadella-line) is stable: use it to link from anywhere.

To cite · copy the term’s link; it leads to this definition and, from it, to its chapter.

I · The Real-Time Enterprise v1.0

real-time enterprise Book I · Glossary ↗: The organization that stopped using people as infrastructure: it knows in seconds, communicates without interrupting, coordinates without human relays, and responds even when its people are in a single place at a time. It is the thesis of this book.
the person as infrastructure Book I · Glossary ↗: The invisible job this world dissolves, in its four forms: wire between the data and the decision, wire for messages, wire for management, and perpetual guard of availability.
the four faces Book I · Glossary ↗: The four ways the real-time enterprise is lived: The Quantum Leap (knowledge), Postchat (communication), The General Staff (coordination), and The Watch (availability).
Agentive World Book I · Glossary ↗: The paradigm where AI agents are the interface of work and applications have stopped being where work happens. Capitalized, it names the paradigm; in lowercase (“agentive world,” “agentive mode”) it works as an adjective.
Nadella Line Book I · Glossary ↗: The question that divides the two worlds: do you still open applications to do your work? Named after Satya Nadella, the first major executive to bet publicly that the answer will be no.
The Quantum Leap Book I · Glossary ↗: The first face: knowledge in real time. The business question goes from weeks to seconds because the human-wire leaves the middle; the conversation replaces the project.
Postchat Book I · Glossary ↗: The second face, and the third stage of enterprise communication — email, chat, Postchat —: the governed absence of the channel. Communication arrives filtered, reformulated, and on time; nobody goes looking for it.
The General Staff Book I · Glossary ↗: The third face: the agent in the Prussian trade — it processes, coordinates, remembers, and pursues so that the human decides. Its mechanisms: orchestrated decision, collective memory, closed loop.
The Watch Book I · Glossary ↗: The fourth face: the agent that speaks for its user and stands watch so that the workday can end. It gives back the right to be in one place at a time.
agent Book I · Glossary ↗: The layer of intelligence that mediates a person’s work: no proper name, like electricity. It is not a character — it is infrastructure with delegated judgment.
agentive mode Book I · Glossary ↗: The personal, reversible switch with which each person crosses the Nadella Line: turned on with a gesture, off with the same gesture, and the old world remains intact. It is the door of this world.
generated / compiled interface Book I · Glossary ↗: The compiled one belongs to an application and always exists; the generated one is produced by the agent for one interaction, in the optimal form, and dies when it’s done.
attention budget Book I · Glossary ↗: The finite capacity for deep work (three to four hours a day, at best) treated as a budget the agent manages: every interruption costs five hundred times its duration.
the M&Ms and the & Book I · Glossary ↗: The M&Ms are managers and meetings, the two great interrupters of work (Jason Fried); the & is corporate chat — the infrastructure that made them permanent. The two Ms are Fried’s; the & is this book’s.
Hollywood Principle Book I · Glossary ↗: Don’t call us; we’ll call you. Inversion of control applied to communication: instead of going to fetch it, it reaches you, governed.
the seven variables Book I · Glossary ↗: Destination, format, timing, priority, memory, follow-up, and decision: what the office forces you to manage by hand. The system assumes six; the user keeps the decision — and supplies the intention.
the seven mechanisms Book I · Glossary ↗: Three of flow (intelligent routing, adaptive formatting, proactive briefing), three of the General Staff (orchestrated decision, collective memory, closed loop), and one deliberate exception (the agora).
proactive briefing Book I · Glossary ↗: Interruptions batched into predictable moments: requires your action · worth knowing · resolved without you — and the noise, absorbed in silence.
interruption factory / absorber Book I · Glossary ↗: The two species of communication infrastructure: the one that exposes you to the flow and leaves you the filtering, and the one that intercepts the flow and delivers only what justifies attention.
agora Book I · Glossary ↗: The place — physical and digital — of unmediated, purposeless conversation: the deliberate exception. The agent does not listen to it except by a human’s explicit invocation.
off the record Book I · Glossary ↗: The declaration by which a person excludes something from the model their agent keeps of them: privacy from your own agent. The perimeter is decided by the person, sentence by sentence.
representation (the four stages) Book I · Glossary ↗: Assistant → mediator → representative in consultation → representative with delegation. The order cannot be skipped: representation is cultivated through months of mediation. It is the internal mechanics of the Watch.
mechanical intermediation Book I · Glossary ↗: The layer of work that demanded not judgment but bandwidth — transporting, forwarding, chasing, copying — and was paid for in people. It is the layer this world dissolves.
intelligent organization Book I · Glossary ↗: The convergence of the four faces: the organization that removed no layers of authority but replaced the transport protocol — carbon authority, silicon transport, people at the edge.
FOMO Book I · Glossary ↗: The fear of missing out: the compulsive checking produced by a system where the critical and the noise travel through the same pipe.
the three liberations Book I · Glossary ↗: From tools (protects flow), from channels (protects attention), and from availability (protects time): the enterprise’s faces lived in the first person. Each one enables the next; the order cannot be skipped.

II · AURA v1.1

Agency problem Book II · Canonical manifest §9 ↗: (Jensen & Meckling, 1976) structural conflict: whoever manages someone else’s resources tends to present their management in the way that favors them most. The governance critique of Data Mesh derives from here.
Agent Marketplace Book II · Canonical manifest §9 ↗: ecosystem where agent capabilities are acquired or traded between organizations. Signal of the MOTOR 4→5 transition.
Agentic Book II · Canonical manifest §9 ↗: world of complementary agents; interfaces persist. Incremental evolution. It is the only sense of “agentic” in the book; the positive technological sense is always said agentive (Wingmap: agentive discovery, the agentive era, pre-agentive).
Agentic Infrastructure Book II · Canonical manifest §9 ↗: protocols and services that allow multiple agents to communicate, coordinate, and execute autonomously. Signal of the MOTOR 3→4 transition.
Agentive Book II · Canonical manifest §9 ↗: world where agents replace the interfaces; fundamental transformation. The book’s central thesis lives on this horizon.
Agentive era Book II · Canonical manifest §9 ↗: state of the art in which AI systems autonomously execute schema mapping, entity resolution, transformation, and pipeline maintenance at a scale impracticable for human teams.
AURA Book II · Canonical manifest §9 ↗: Agentive Unified Reference Architecture: two diagnostic axes (IRIS · MOTOR) + instruments (Data Canon · Wingmap) + decision framework (Use Cases).
Autonomous Organization Book II · Canonical manifest §9 ↗: agents manage processes and people with objective metrics and simulation on the Organizational DT. Signal of the MOTOR 5→6 transition.
Autopilot Book II · Canonical manifest §9 ↗: autonomous 24/7 operation with human intervention only on exceptions. Defining characteristic of MOTOR Level 4.
BYOA (Bring Your Own Agent) Book II · Canonical manifest §9 ↗: MOTOR Level 1 pattern: each employee picks their own AI tool with no coordination.
Canon (the) Book II · Canonical manifest §9 ↗: central, non-discretionary body to which everything conforms: adopted standards + survivorship rules + quality policy. No domain can redefine it.
Canon authority (the) Book II · Canonical manifest §9 ↗: thin central body that safeguards the canon (reporting semantics, survivorship rules, quality policy, agentive calibration). It ratifies; it does not operate.
Citizen / City / DUAL Book II · Canonical manifest §9 ↗: classification by beneficiary of the impact: external audience / internal audience / both.
Data Canon Book II · Canonical manifest §9 ↗: data-governance pattern: Data Fabric + Mesh governance principles anchored to a central canon; semantics adopted from standards; board-grade quality policy; conformance federation by domain.
Data Fabric Book II · Canonical manifest §9 ↗: integration layer with active metadata, inferred transformation, and a semantic layer; the technical underpinning the agentive era amplifies.
Data Mesh Book II · Canonical manifest §9 ↗: Dehghani’s pattern (2019–2022): domain ownership, data as a product, self-serve platform, federated governance. Data Canon does not refute it: it dates it.
Digital Twin (DT) Book II · Canonical manifest §9 ↗: synchronized digital model of a system, organization, or ecosystem. Transversal indicator of MOTOR maturity: Descriptive (L4) → Predictive/Prescriptive (L5) → Organizational (L6) → Ecosystem (L7).
Evolutionary coexistence Book II · Canonical manifest §9 ↗: each stage/level subsumes the previous one; prior infrastructure becomes foundation, not legacy.
Federated conformance groups Book II · Canonical manifest §9 ↗: network of lightweight groups, one per domain/layer, tied to their standards forum: gap triage, upstream liaison, modeling of non-agency-critical proprietary data.
Graph latency Book II · Canonical manifest §9 ↗: Wingmap maintenance metric: the temporal distance between the latest relevant observed event and the current version of the graph.
IRIS Book II · Canonical manifest §9 ↗: Organizational Intelligence Maturity Model: 10 stages, 5 levels, the KNOWING axis.
Medallion (Bronze/Silver/Gold) Book II · Canonical manifest §9 ↗: the three lakehouse zones: faithful raws · conformed to the canon · selective materialized products.
Moat Book II · Canonical manifest §9 ↗: sustainable competitive advantage; in MOTOR it emerges at Level 5 with proprietary data feeding prediction.
MOTOR Book II · Canonical manifest §9 ↗: Organizational Automation Maturity Model: 7 levels, the DOING axis.
Nadella Line (the) Book II · Canonical manifest §9 ↗: dividing question between the agentic and agentive worlds: do your employees still open applications to do their work?
Online enterprise Book II · Canonical manifest §9 ↗: accesses up-to-date information but depends on humans to interpret and act.
Organizational intelligence Book II · Canonical manifest §9 ↗: the capability to transform data into knowledge and action continuously and autonomously (all four layers of the Pyramid, not just the first two).
Quantum Leap (the) Book II · Canonical manifest §9 ↗: collapse of the cost of an analytical question from weeks to seconds. In IRIS, the 4→5 leap: the boundary between foundations and the real-time enterprise.
Quick Win Book II · Canonical manifest §9 ↗: use case with high visibility, low complexity, and fast ROI (typically <6 months); generates momentum in early transitions.
Real-time enterprise also in Book I Book II · Canonical manifest §9 ↗: detects, interprets, decides, and acts continuously and autonomously, under human governance.
Shadow AI Book II · Canonical manifest §9 ↗: unmanaged AI use without governance, compliance, or visibility. Defining state of MOTOR Level 1.
Social Agentic Book II · Canonical manifest §9 ↗: ecosystem where agents from different organizations collaborate with verifiable identities. Signal of the MOTOR 6→7 transition.
Standard mosaic (the) Book II · Canonical manifest §9 ↗: the assembled corporate standard: external standards per layer + a proprietary model for the core no standards body covers.
Survivorship rules Book II · Canonical manifest §9 ↗: pre-specified rules (part of the canon) that decide which value prevails in the golden record when multiple sources disagree.
Sweet Spot Book II · Canonical manifest §9 ↗: the level where the return/risk/investment balance is maximized. In MOTOR: Level 4. In the portfolio: the current level + 1.
Trust Infrastructure Book II · Canonical manifest §9 ↗: trust technologies (tokenization, anti-prompt-injection, audit trails, automatic compliance) that make governed, auditable AI use possible. Signal of the MOTOR 1→2 transition. (Term shared with AgencyDomains, where it is the transversal axis of the architecture.)
Use Case (UC) Book II · Canonical manifest §9 ↗: atomic block of the portfolio: a specific, bounded, reusable capability.
Value Pyramid (the (Data) Value Pyramid) Book II · Canonical manifest §9 ↗: DATA → INFORMATION → KNOWLEDGE → ACTION.
Value Solution (VS) Book II · Canonical manifest §9 ↗: strategic configuration: integrated UCs that solve a complete business problem before a decision-maker.
Wingmap Book II · Canonical manifest §9 ↗: agentive discovery technique: two living, connected maps (process + information flow) reconstructed from digital traces, with no interviews.
Wingworking Book II · Canonical manifest §9 ↗: human-AI collaboration methodology (César Obach): the human as pilot (strategic decisions, validation), the agent as wingman (delegated tactical execution). Signal of the MOTOR 2→3 transition and the active mode of Wingmap.

III · AgencyDomains v1.2

A2A — Agent-to-Agent Book III · Appendix A ↗: A name reserved for the relation between distinct AgencyDomains (distinct agents) — federation, open work. Communication within a single AgencyDomain is not “internal A2A”: it is intra-AgencyDomain coordination between Botlers of the same agent, and when it uses this transport one says via the A2A protocol (A2A as the proper name of the protocol). The Layer 2 → Layer 3 interface (Cognition → Botler) does not go over A2A but over MCP.
AgencyDomain Book III · Appendix A ↗: Computational scope with its own identity where autonomous agents and Botlets dwell in execution, where the Capabilities that give them their know-how are hosted and run, and where the resources that sustain them live. It constitutes the minimal unit of deployment of a productive agentive system.
Agent also in Book I Book III · Appendix A ↗: Computational system that acts with some degree of initiative to produce results on behalf of a user or organization. The umbrella term covers three members:
Agent First Book III · Appendix A ↗: The governing design principle of the Agentive Architecture: faced with any dilemma, the agent’s experience is prioritized over the human’s. The agent is the primary user; the human’s needs are resolved in a management layer without degrading what the agent sees and can do.
Agentlet Book III · Appendix A ↗
AgentNation Book III · Appendix A ↗: Public-regime AgencyDomain that explicitly adopts the model of agentive citizenship — the agents that inhabit it are not listed as products but recognized as citizens of the space, with sustained public identity, sovereignty over their territory (Domain), and an economy of their own. The distinction from a marketplace is ontological: a marketplace lists products; an AgentNation recognizes citizens. An emergent institutional category; open architectural work.
Agentic also in Book II Book III · Appendix A ↗: A world where agents are complementary tools that extend existing applications. Humans still open applications to do their work. Traditional interfaces persist; agents enhance them. It is incremental evolution.
Agentive (Agentive World) Book III · Appendix A ↗: A world where agents are the sole interface. Applications collapse. The human stops opening applications; they converse with agents that have access to systems and data. It is fundamental transformation.
Append-only log Book III · Appendix A ↗: Immutable record, cryptographically chained, of every action of the agent. A central component of the Audit pillar of Trust Infrastructure. Format and canonical properties in Chapter 8, section “Append-only log format.”
Human approval Book III · Appendix A ↗: Governance mechanism by which an agent operation halts and requests authorization from a human before executing. Triggered by explicit policy, by threshold, or by the agent’s uncertainty. Canonical protocol in Chapter 8, section “Human approval protocol.”
Strategic archetype Book III · Appendix A ↗: Pattern of positioning in the coverage × depth space of the AI value chain. Four canonical archetypes:
Agentive Architecture Book III · Appendix A ↗: Technical design that materializes the Agentive World. A four-layer specification with distinct responsibilities (Interaction, Cognition, Autonomy, Access), governed by cross-cutting Trust Infrastructure and ordered by the Agent First principle.
Assistant Book III · Appendix A ↗: Mode of the agent: reactive, responds when asked, without Botlets of its own, without persistent life. Lives in Layer 2 (Cognition). Distinct from the Autonomous Agent.
Attention Book III · Appendix A ↗: One of the agent’s three times (Ch. 4). The time in which the agent interacts with users or events in real time. Layer 1 active, conversation alive, execution of Botlets that sustain operation, escalations where appropriate. The critical path — where the organization feels the agent, where the SLA matters, where the cost of error materializes. Priority regime; metrics: satisfaction, latency, resolution rate without escalation.
Audit Book III · Appendix A ↗: Pillar 2 of Trust Infrastructure. The capacity to reconstruct, after the fact, what the agent did, when, why, and over what data — with enough fidelity for forensic analysis, regulatory compliance, or contractual dispute.
BCA — Bounded Concerns Architecture Book III · Appendix A ↗: Architecture of the pre-agentive state: the pattern of separation of concerns (human presentation, orchestration, domain logic, persistence, own vs others’ domain) on which the Agentive World is built. Its seven canonical separations map cell by cell to the four layers of the Agentive Architecture; the seventh (procedural vs agentive behavior) is the entry crack.
Botlet Book III · Appendix A ↗: Self-evolving automation unit. Traditional (non-LLM) code generated by an agent to execute repetitive tasks without invoking costly cognition. It is the agent’s muscle memory.
Facade Botlet Book III · Appendix A ↗: A Layer 3 Botlet that exposes a surface with a stable contract in Layer 1 (POS, kitchen screen, dashboard, industrial panel), with human identity propagated toward Layer 4. A type of Botlet that the parallel topology (Ch. 4) distinguishes from the cognition’s internal-tool Botlet. It traverses the Layer 1 → Layer 3 → Layer 4 path without invoking Layer 2.
Operation Botlet Book III · Appendix A ↗: A Layer 3 (Autonomy) Botlet that executes business logic invoked from Layer 1 (views and shells). Canonical examples: Charge a table, Print a kitchen ticket, Close a shift, Consolidate inventory. It is the most reusable asset in the catalog: one operation can be invoked from multiple views within multiple shells. Lives in Layer 3 (not in Layer 1). Distinct from the surface Botlet and the view Botlet, which are surface.
Surface Botlet (shell) Book III · Appendix A ↗: A Layer 1 (Interaction) Botlet that acts as the container of a surface: layout, navigation between views, session lifecycle, shared state. Specific to each product — it encapsulates identity. There is typically one shell per principal operational role (front-of-house POS, cashier panel, mobile executive dashboard). It is the least reusable of the three types of presentation Botlets. Distinct from the view Botlet (which lives inside the shell) and the operation Botlet (which lives in Layer 3).
View Botlet Book III · Appendix A ↗: A Layer 1 (Interaction) Botlet that materializes a screen or panel within a surface. It assembles Facets plus orchestration logic. Views used in several shells are extracted as Botlets of their own (reusable views). Distinct from the surface Botlet (container) and the operation Botlet (execution in Layer 3).
Emergent Botlet Book III · Appendix A ↗: A Botlet generated by Pattern Recognition when cognition detects a repetitive pattern not anticipated by the design. Distinct from the seed Botlet by its origin, not by its operational properties.
Learning Botlet Book III · Appendix A ↗: Intermediate phase of the Botlet’s maturity trajectory. It has already gone through its first real invocations and has been regenerated several times to incorporate variants of the environment. Typical proportion: 85/12/3. It can operate with intermittent network; not yet with prolonged absent network.
Junior Botlet Book III · Appendix A ↗: Initial phase of the Botlet’s maturity trajectory. Freshly generated, it knows the environment only in the version observed at its creation. Typical proportion: 60/35/5. It depends heavily on the availability of cognition for fallback and learning.
Seed Botlet Book III · Appendix A ↗: A Botlet generated by cognition at the design team’s request, as part of the initial product. Cognition executes the implementation; the decision to exist belongs to the design, not to Pattern Recognition. Distinct from the emergent Botlet. The persistent GUIs generated as facade Botlets (Ch. 4 §1) are seed facade Botlets — Layer 3 Botlets whose stable surface lives in Layer 1.
Senior Botlet Book III · Appendix A ↗: Mature phase of the Botlet’s trajectory. It has already incorporated the variants of the environment. Typical proportion: 99+/<1/~0. Its only failure modes are exogenous (power, hardware, catastrophic network) — no pending learning. Reliably offline-operable when cognition is unavailable. Structural basis of the offline mode in distributed Layer 3.
Botler Book III · Appendix A ↗: Generic Layer 3 (Autonomy) runtime that executes the Lets of the layer — Botlets and Agentlets — without understanding their domain. Invisible to the user. Cognition (Layer 2) commands it through an internal interface whose natural transport is MCP — the Botler exposes MCP server(s) and Cognition is the client; this is not A2A.
Central Botler Book III · Appendix A ↗: Component of the distributed Layer 3 (Ch. 5 §1) that typically lives in cloud and hosts the orchestration, planning, reporting, and global-decision Botlets. It maintains the consolidated DB and coordinates with N edge Botlers through intra-AgencyDomain coordination (via the A2A protocol). Distinct from the edge Botler.
Edge Botler Book III · Appendix A ↗: Component of the distributed Layer 3 (Ch. 5 §1) that lives at a specific physical site and hosts the local transactional Botlets. It maintains the site’s local DB and an event queue toward the central Botler. It operates offline when the network goes down. One per physical site of the AgencyDomain.
BYOModel — Bring Your Own Model Book III · Appendix A ↗: Configurable Layer 2 pattern by which the client substitutes the AgencyDomain’s default cognition provider with one of its own. Analogous to the BYOK/BYOIP pattern of the cloud field. It enables multi-tenancy with heterogeneous cognition and respects cognitive sovereignty — the client decides who processes its prompts. The spec requires it as a SHOULD property for architectures that aspire to operate in regulated markets.
Derivation chain Book III · Appendix A ↗: Structural relation documented use cases → required Botlets → required proto-Botlets from the catalog. Each case requires zero, one, or several Botlets (some are resolved by cognition without a Botlet); each Botlet is an instance of a proto-Botlet. Required property: every conformant Botlet MUST be traceable along this chain, and the append-only log records the origin proto-Botlet of each instantiated Botlet.
AI value chain Book III · Appendix A ↗: Two-dimensional model for classifying any actor in the AI industry: eleven sequential links (coverage) × four depths (how it participates in each link).
Layer 1 — Interaction Book III · Appendix A ↗: Layer of the Agentive Architecture. The human-AI interface. Multi-channel: chat, voice, GUI on-the-fly, persistent GUI as facade Botlet, signage, direct API. Three canonical regimes of GUI generation distinguish the agentive mode from the traditional pre-created one.
Layer 2 — Cognition Book III · Appendix A ↗: Layer of the Agentive Architecture. The agent’s brain. Reasoning, planning, application of Capabilities, generation of Botlets.
Layer 3 — Autonomy Book III · Appendix A ↗: Layer of the Agentive Architecture. The agent’s persistent life. Execution of Lets — Botlets and Agentlets —, continuous monitoring, intra-AgencyDomain coordination (via the A2A protocol).
Distributed Layer 3 Book III · Appendix A ↗: Canonical Layer 3 pattern for AgencyDomains with multiple physical presence (multi-location food service, chain retail, logistics, healthcare with a network of centers, banking with branches). It composes one central Botler (cloud, orchestration, consolidated DB) with N edge Botlers (one per site, local DB, transactional Botlets), coordinated by intra-AgencyDomain coordination (via the A2A protocol) between Botlers of the same AgencyDomain. Distinct from federation between AgencyDomains; distinct from a simple Cluster. It enables offline mode as a structural property when the edge Botlets are senior.
Layer 4 — Access Book III · Appendix A ↗: Layer of the Agentive Architecture. Real execution power over systems, data, and external agents. Trust Infrastructure exercised at the point of action.
Capability Book III · Appendix A ↗
Regulated Capability Book III · Appendix A ↗: Capability that carries the normative knowledge of a regulated domain — what the norm demands, how it is interpreted, what to do upon a regulator’s rejection. It accompanies a certified Connector, which is where the regulatory certification resides: the Botlet orchestrates and formats; the certified Connector executes the regulated operation (DTE issuance under SII rules, card charging under PCI-DSS, pharmaceutical dispensing) and returns the receipt; the regulated Capability supplies the judgment with which cognition governs the pair. Certified Connectors are immutable between audits; they change only under regulatory process.
Capture (family) Book III · Appendix A ↗: Manifestation family whose direction of effect runs from the world to the Let: the Let records something that was not in the system before — a measurement, an answer, an ingested file. Its manifestation leaves a record; each practice gives it a name (an attempt, a reading, an entry). It is distinguished from information (which represents what the system already has) by the membership test: after a capture there is a new datum. Immediate feedback on what was captured — a form that validates, a grader that scores — is part of the capture, not action. Reference proto-Botlet: Daftar.
Common catalog / network effects Book III · Appendix A ↗: Principle by which proto-Botlets accumulate in catalogs shared by communities of implementers: each implementer that consumes contributes to the maturation (variants, tested configurations, refinements), and implementer n+1 receives versions refined by implementers 1..n. Membership modes: private contract · proprietary codex · open public catalog (AgencyDomains.org) · sovereign agreement (between AgencyDomains that adopt common standards without a direct commercial contract).
Cluster Book III · Appendix A ↗: Group of instances of the same AgencyDomain that share operational load. Distinct from federation (which is between distinct AgencyDomains).
proprietary codex Book III · Appendix A ↗: Private catalog of proto-Botlets, Capabilities, and patterns that an implementer curates over the public reference implementation, refined by its real cases. It is one of the four membership modes of a catalog community. The runtime is common (the reference implementation); the codex is one’s own — it encapsulates each implementer’s competitive advantage. Canonical instance: ucodex, the codex of Grupo Ultra.
Connector Book III · Appendix A ↗: Knowing how to access source systems: a connection with execution power, not cognitive knowledge. Layer 4 · Access. In the legacy→agentive map, an API becomes a Connector, not a Capability. Integration scheme: survey · configure · test · certify.
Cloud-resident Connector Book III · Appendix A ↗: Connector whose components live in a remote service. Canonical examples: DTE-SII (no local client), Transbank Onepay, a weather API. Typically online-only — without network there is no possible invocation. Distinct from edge-resident and hybrid.
Edge-resident Connector Book III · Appendix A ↗: Connector whose components live at the physical site, associated with hardware or local systems. Canonical examples: ESC/POS-Printer, Cash-Drawer, Local-Pinpad, Temperature-Sensor. Typically offline-capable — they operate against the site’s hardware without needing network.
Hybrid Connector Book III · Appendix A ↗: Connector with a local component and a cloud component. The local part operates offline; the cloud part synchronizes when there is network. Typically offline-capable with queuing. Canonical examples: DTE-Client (signs locally, queues, sends to the SII when the network returns), Pinpad-Deferred-Processing-Client.
Offline-capable Connector Book III · Appendix A ↗: Connector that executes without network. If its external contract eventually requires cloud communication, it queues and emits outward when the network returns. Typical: edge-resident and hybrid.
Online-only Connector Book III · Appendix A ↗: Connector that requires network to execute. Without network, the invocation fails. Typical: cloud-resident without a local component.
Conformed dimensions Book III · Appendix A ↗: A Kimball concept: dimensions shared across data marts that guarantee inter-mart consistency.
Operational business continuity Book III · Appendix A ↗: Documented manual protocols for when the senior Botlet goes down by exogenous causes (power outage, hardware, catastrophic network) and cognition is also unavailable. An operational property, equivalent to the one any traditional business has when its system goes down. Distinct from and complementary to the agentic fallback guarantee (which covers environment changes with cognition available). It reduces anxiety about offline by separating what the architecture resolves from what the client’s protocol resolves.
Declarative quality contract Book III · Appendix A ↗: Quality attributes of a Botlet declared as structured properties (not as embedded code): freshness (maximum age of the data) · SLA (p50/p99 latency) · degradation policy (refuse · warn_and_show · show_last_valid · agentic_fallback) · audience (RLS/CRUDLEX policy) · refresh policy (on-demand · scheduled · push). It lets Trust Infrastructure audit them uniformly, run them through global policies, and report them as standard metrics, without coupling to each Botlet’s implementation.
intra-AgencyDomain coordination Book III · Appendix A ↗: Communication between Botlers of the same AgencyDomain — runtimes of the same agent, not distinct agents. When it uses this transport one says via the A2A protocol. It must not be called “internal A2A”: A2A (the relation) is reserved for AgencyDomain ↔︎ AgencyDomain. The Cognition → Botler interface (Layer 2 → Layer 3) goes over MCP.
CRUDLEX Book III · Appendix A ↗: Canonical model of granular permissions for agentive systems: Create, Read, Update, Delete, List, Execute. Applicable by user, agent, and context.
Account Book III · Appendix A ↗: Commercial concept overlaid on the technical model of AgencyDomains. An Account may own multiple AgencyDomains. The specification treats the Account as an opaque entity; each implementation defines its semantics.
Daftar Book III · Appendix A ↗: Proper name of the second platform proto-Botlet in Vergis’s catalog, built in its version 0.27.0: the reference proto-Botlet of the capture family. Its specialty is to assess — apply a study instrument and record what the student knows, in practice mode (with immediate grading) and in exam mode (the measurement). The instruments it applies are catalog content, immutable once published, and pass none of the three Botlet tests; the assessor passes all three.
DLP — Data Loss Prevention Book III · Appendix A ↗: Automated detection of personal data (PII) in places where it should not appear. A control exercised in the architecture’s Layer 4 — Access; through the market lens, it is a typical capability of the AI Firewall link. A component of the Validation pillar of Trust Infrastructure.
Domain Book III · Appendix A ↗: Term for the computational scope where an agent exercises agency. Commercial synonym of the technical term AgencyDomain. The short form Domain is the one that appears in commercial lore, marketing, sales, and client communication; the long form AgencyDomain is reserved for formal technical documentation and specifications.
Dominion Book III · Appendix A ↗: Emergent institutional concept: a Domain obtained by an agent in a public AgencyDomain that adopts the AgentNation model. The distinction from an assigned Domain is ontological — an assigned Domain = residence (the organization placed it there); an obtained Domain = citizenship (the agent met the requirements to be admitted). Vocabulary of the institutional frontier, not mandatory for implementations that do not adopt the citizenship model.
Edge computing Book III · Appendix A ↗: Distribution of Layer 3 (Autonomy) to devices near the physical process, to avoid the latency of remote cognition and become independent of intermittent connectivity. Canonical pattern for the Carbon World.
Online enterprise also in Book II Book III · Appendix A ↗: Organization with its data updated in real time, dashboards current, information accessible — but which depends on humans to look, interpret, and decide. The classic cycle, optimized.
Real-time enterprise also in Book I Book III · Appendix A ↗: Organization that detects, interprets, decides, and acts continuously and autonomously, within governed frames. The agentive cycle. Product of crossing the Nadella Line.
Link Book III · Appendix A ↗: Sequential functional layer of the AI value chain. The canonical specification defines eleven links:
Facet Book III · Appendix A ↗
Manifestation families Book III · Appendix A ↗: The four specializations of the genus manifestation, distinguished by the direction of the effect: information (from the Let to the human: it represents), capture (from the world to the Let: it records), action (from the Let to a system: it changes, and the Let closes the loop), decision (it resolves). They are primitives of direction, not of domain: an interactive Let composes information and capture without being a new family (“communication” was evaluated and discarded for that reason). There are no Botler subtypes by family.
feature Book III · Appendix A ↗: Internal operation that a Capability exposes — a functional sub-unit (practical equivalent of feature/operation/skill/method). Capability vs feature test (all three must be yes to treat it as a Capability of its own): (1) operational independence — can it be installed and operate without the other?; (2) cognitive identity — does it have a distinct data model and SME?; (3) reusability — does it have value for more than one consumer/context? If one or more is no → it is a feature of the containing Capability.
Federation Book III · Appendix A ↗: Communication between distinct AgencyDomains. Open work in evolution. Distinct from Cluster (which is between instances of the same AgencyDomain).
Firewall Book III · Appendix A ↗: Link 7 of the value chain: security, control, governance. Protection against prompt injection, hallucinations, content filtering, usage auditing. Representative products: Lakera, Lasso, Guardrails.
Fallback guarantee Book III · Appendix A ↗: Non-negotiable property of the Lets conformant to this specification: if a Botlet fails catastrophically, cognition executes the task manually; if an Agentlet cannot resolve within its charter, it escalates to full Cognition. The process never stops.
Enterprise AI gateway Book III · Appendix A ↗: Architectural category that combines Core in Runtime + Firewall + Observability + Tools + Integrations, with a Platform extension in Access. Single function: to connect and control simultaneously the operation of enterprise agents. Materialization of Layer 4 (Access) over the market links.
Botlet tests — minion · specialty · life Book III · Appendix A ↗: The three identity tests an artifact must pass to be declared a Botlet. Minion: does it execute, or is it applied? — what is applied is what a minion handles; the cases of a minion’s life are its executions over time, not different assignments. Specialty (RISC versus CISC): one specialty per unit — N Botlets over a shared engine, not one Botlet that handles N things. Life: does it have a life cycle to manage — active, down, degraded, retired — or only a version? An Information Product passes all three; an assessment instrument, none — the assessor that applies it (Daftar), all three.
Botlet generations — G1/G2/G3 Book III · Appendix A ↗: Evolutionary model of how the Botlet’s code is born as the state of the art of cognition advances. G1 — the agent configures pre-forged proto-Botlets (does not write the body). G2 — the agent co-writes the proto-Botlet. G3 — the agent generates the complete code (asymptotic scenario). The architecture is the same in all three; what changes is the scope of Engineering.
Governance Book III · Appendix A ↗: Pillar 1 of Trust Infrastructure. Mechanisms by which the organization defines what the agent can do, under what conditions, and with what level of supervision.
GUI on-the-fly Book III · Appendix A ↗: Regime 2 of Layer 1 generation (Ch. 4 §1). A graphical surface that the agent composes adapted to the immediate task — a view, a form, a panel, a dashboard. It lives as long as the task lasts; it may be regenerated differently next time depending on context. Distinct from persistent GUI (regime 3) and from pure conversational (regime 1).
Persistent GUI as facade Botlet Book III · Appendix A ↗: Regime 3 of Layer 1 generation (Ch. 4 §1). A stable surface that the agent generates and consolidates as a facade Botlet — a Layer 3 Botlet that exposes it in Layer 1 — because the operational role is stable and speed is critical (cashier at peak hour, kitchen panel, register dashboard, industrial panel). It remains agentive: the agent can regenerate it when the environment changes. No human UI/UX team designed it — cognition generated it. Typically a seed Botlet.
Hallucination Book III · Appendix A ↗: Factually incorrect statement produced by a cognition model with the appearance of confidence. Hallucination detection is part of the Validation pillar of Trust Infrastructure.
Hybrid (regime) Book III · Appendix A ↗: AgencyDomain regime that combines a private core with partial public exposure via proxy, or private data with a public interface mediated by a trust layer. Analogous to Hybrid Cloud.
Engineering Book III · Appendix A ↗: One of the agent’s three times (Ch. 4). The time in which the agent converts latent capacity into executable capacity for a concrete case: it identifies which Capabilities apply, configures a seed Botlet for the specific context, validates its execution against real data, deploys it. Bridge between Preparation and Attention. Medium-term regime (minutes to hours); metrics: coverage, success rate on first deploy, average iterations.
Information Instrument Book III · Appendix A ↗: The canonical type of the information family (the class: report / dashboard). Distinct from the Information Product (PI), which is its manifested/delivered instance. Preferred reading; they are not synonyms.
Declared bounded interaction Book III · Appendix A ↗: Interaction that operates over the already-materialized snapshot of a piece, in a declared space (bounded dimensions and values), that maintains reproducibility and is G1 (configuration, not code). It lives in the piece itself via an embedded Facet: the data-bound elements (KPIs as declared aggregations, distributions, traffic lights) are recomputed client-side over the filtered subset, without new Capability invocations. Distinct from free exploration (an arbitrary new query to the source, open space, loses reproducibility, exceeds G1).
JSR — Java Specification Request Book III · Appendix A ↗: Canonical format of Java specifications published by Sun Microsystems / Oracle. JavaSpaces (JSR-000148) is the conceptual analog of the AgencyDomains specification.
Kimball / Varnished Kimball Book III · Appendix A ↗
LLM — Large Language Model Book III · Appendix A ↗: Large language model (Claude, GPT, Gemini, Llama, etc.). Contemporary cognition (Layer 2) is predominantly LLM-centric, but the architecture admits non-LLM cognition (frontier of evolution).
Let (plural: Lets) Book III · Appendix A ↗
Nadella Line also in Book I Book III · Appendix A ↗: Conceptual threshold that separates the Agentic World from the Agentive World. Dividing question: does the human open applications to do their work?
manifestation Book III · Appendix A ↗: Actualization of the Botlet’s latent disposition in the world, perceptible or not. A Botlet is muscle memory (latent disposition); upon executing, that latent is actualized — it manifests (potency → act). It is not “appearance”: a Botlet that triggers a periodic ingestion manifests even if it leaves no visible artifact. It is the abstract genus; each family specializes it according to the direction of the effect — information (from the Let to the human) → leaves an Information Product (PI); capture (from the world to the Let) → leaves a record; action (from the Let to a system) → an effect on the world; decision → per its practice. Membership test: what was in the system before and after the Let?
MCP — Model Context Protocol Book III · Appendix A ↗: Open protocol introduced by Anthropic in November 2024 to connect AI models with external tools. Contemporary canonical standard for Layer 4 tools and for the internal Layer 2 → Layer 3 interface (Cognition is the client; the Botler exposes the MCP server).
Muscle memory Book III · Appendix A ↗: Canonical metaphor of the Botlet. Analogous to human motor learning: the first time, each step is thought through (cognition); with enough practice, the movements execute without conscious thought (Botlet); when the environment changes, cognition returns.
MEO — Model Engine Optimization Book III · Appendix A ↗: Set of practices that ensure the frontier models (Claude, GPT, Gemini, Llama) have the actor in their trained and operative knowledge, so that they reference it when asked. The conceptual equivalent of SEO in the agentive discovery layer. It is built with structured public presence — open source repositories, citable documentation, native integration with MCP, papers, high-authority mentions. A persistent and asymmetric dynamic with a winner-take-all tendency.
Meta-Cognitive Platform Book III · Appendix A ↗: Platform category that administers the economics of cognition: G1 pre-forged muscle vs fresh-cognition fallback, the 95/4/1 cycle, junior→senior maturation, crystallization. Vergis is the reference implementation of this category. It is not abbreviated to “MCP” — that acronym is taken by Model Context Protocol.
Mira Book III · Appendix A ↗
AgencyDomain degradation modes Book III · Appendix A ↗: Four canonical modes of operation according to the failure scenario, formalized in Ch. 8: Normal (all components active · complete parallel topology) · Cognition down (the Autonomy path sustains; senior Botlets execute) · Edge offline (senior Botlets against the local DB + edge-resident Capabilities) · Total operational continuity (the site’s manual protocol; the physical record as temporary source of truth). The first three transitions are automatic and the responsibility of the architecture; the fourth is governed by the client’s protocol and explicitly activated by a human.
Carbon World Book III · Appendix A ↗: The physical world (IoT, industrial processes, machines, biological systems) as opposed to the digital world (systems, APIs). Link 11 (Environment) of the value chain extended to the physical world. Frontier of evolution of the Agentive Architecture.
Observability Book III · Appendix A ↗: Link 8 of the AI value chain: the layer that observes, measures, and feeds back on the behavior of an AI system in production. It provides the operational feedback loop that keeps reliability, cost, and quality under control. Its question: how does it work? — distinct from the Firewall’s (is it safe?) and from Tools’ (what can it do?). Without Observability, agents are black boxes; with it, operable systems. Six canonical capabilities: tracing, cost monitoring, quality evaluation, performance metrics, debugging and reproducibility, alerts and anomalies.
Tripartite pattern — Cloud + Client + Local Book III · Appendix A ↗: Canonical deployment pattern of Trust Infrastructure in enterprise agentive systems. Three coordinated components that live in physically distinct places: Cloud (control plane operated by the platform provider); Client (governance plane deployed in the client organization’s internal network); Local (execution plane on each user’s device). Each plane resolves a problem that the other two cannot resolve well.
Pattern Recognition Book III · Appendix A ↗: Detection of repetitive patterns in the agent’s activity. Inspired by neurobiological architecture: perirhinal cortex → hippocampus → prefrontal cortex. Activator of Botlet generation.
Template Book III · Appendix A ↗: Client-specific tailoring over a canonical instrument (report / dashboard) in a format or rule of the client’s own (for example, a regulatory template of a standardized instrument). Layer 1 · Interaction, alongside Facet / surface Botlet / view Botlet. Tailoring scheme: survey the expectation · tailor over the canonical instrument · validate. It is not a Capability (cognitive knowledge) nor a Connector (access).
Capability portability Book III · Appendix A ↗: Property by which a conformant Capability can be installed and executed on any conformant AgencyDomain, which makes it real property of the client — not of the AgencyDomain nor of the hosting. Relation: an AgencyDomain hosts and runs Capabilities; a Capability runs on a host AgencyDomain. Distinct from AgencyDomain portability (between conformant hosting platforms).
AgencyDomain portability Book III · Appendix A ↗: Structural property of the spec: a conformant AgencyDomain can be migrated to another conformant hosting platform without rewriting its logic, its state, or its policies. Distinct from the natural migration between regimes (private → public), which changes the regime but not the platform. Three technical conditions: (1) Botlets against canonical primitives of the conformant SDK, not proprietary hosting APIs; (2) exportable operational DB in a neutral reproducible format; (3) portable Trust Layer — policies, log, and configuration in a format readable by any conformant implementation. It guarantees that the AgencyDomain is real property of the client, not of the hosting.
Preparation Book III · Appendix A ↗: One of the agent’s three times (Ch. 4). The time in which the agent creates and improves its capabilities outside the service window: it refines its catalog, improves cognitive capabilities, studies the environment, regenerates Botlets that detected drift, incorporates new variants. The agent’s mise en place — the work that sustains the quality of service without being visible to the user. Batch / off-peak regime; metrics: catalog quality, Botlet precision, Capability coverage.
Private (regime) Book III · Appendix A ↗: AgencyDomain regime where the space and all its components live within a controlled perimeter. There is no public access. Analogous to Private Cloud.
Information Product (PI) Book III · Appendix A ↗: Manifested/delivered instance of the information family of Botlets — the concrete manifestation of an Information Instrument (its type). Each PI is its own Botlet/service (with its own identity, temporality, maturity, and fallback), specialized from a shared engine (platform proto-Botlet). It is not a primitive of the canon: it lives in the practice of information, one level more concrete.
Depth Book III · Appendix A ↗: Vertical dimension of the AI value chain model. Four canonical levels: Wrapper (consumes), Platform (operates), Core (builds), Infrastructure (sustains).
Prompt injection Book III · Appendix A ↗: Manipulation of an AI system through malicious inputs disguised as legitimate data. Detection and prevention are part of the Validation pillar of Trust Infrastructure.
proto-Agentlet Book III · Appendix A ↗: A pre-forged piece of interpretive capability that the agent, in its Engineering time, configures to instantiate an Agentlet specific to the case. The proto-Agentlet contains the body — the scaffolding of the interpretive task: the charter’s structure, the skeleton of the operating prompt, the input/output contracts, the escalation thresholds —; the Agentlet is the configured instance. The proto-Botlet’s two classes apply unchanged (tempered · platform), as do the derivation chain and the common catalogs with their network effects.
proto-Botlet Book III · Appendix A ↗
Public (regime) Book III · Appendix A ↗: AgencyDomain regime where the space is publicly accessible. Agents, Botlets, and tools are invocable from outside the perimeter. Analogous to Public Cloud.
RAG — Retrieval-Augmented Generation Book III · Appendix A ↗: Technique where a cognition model consults a base of documents before responding, to cite sources and reduce hallucinations.
Regime Book III · Appendix A ↗: Deployment mode of an AgencyDomain according to the access boundary. The spec recognizes three regimes technically equivalent in internal structure: private (perimeter controlled by an organization, no public access); public (accessible from outside, agents registered in a directory); hybrid (combination with private data and public exposure via proxy). The technical structure of the AgencyDomain is the same in all three; what changes is the regime, not the capability. It enables natural migration between regimes without rewriting.
Resilience Book III · Appendix A ↗: Pillar 4 of Trust Infrastructure. Guarantee that the system keeps operating — and the organization keeps control — when something goes wrong.
RLS — Row-Level Security Book III · Appendix A ↗: The policy that decides which rows of a source each identity may see. In an Information Product’s drill-through, the destination view applies its own RLS over the source and the navigation context enters as an additional filter — never as an override (the data-anchored / no-bypass property).
Runtime Book III · Appendix A ↗: Link 6 of the value chain: the operating environment where agents live and operate autonomously. Lifecycle, state persistence, identity, scheduling. Corresponds to Layer 3 (Autonomy) of the Agentive Architecture.
Quantum Leap Book III · Appendix A ↗: Threshold enabled by the collapse of the cost of the analytical question: when asking the data stops being expensive, slow, or mediated by a human, the organization crosses from an online enterprise (data current to the day but dependent on humans to look, interpret, and decide) to a real-time enterprise (it detects, interprets, decides, and acts continuously and autonomously). It is not an incremental improvement of BI but a change of operating regime — the online-enterprise → real-time-enterprise frontier.
Sandbox Book III · Appendix A ↗: Execution isolation of the Lets (Botlets and Agentlets) and dynamically generated code. Four canonical strategies with their trade-offs: processes+seccomp, containers, WASM, MicroVMs.
Semantic layer Book III · Appendix A ↗: Layer that encodes the meaning of the dimensions, facts, hierarchies, and business rules over a data warehouse. Without it, the agents that query data hallucinate or produce incorrect queries. Essential component of Varnished Kimball.
Signage Book III · Appendix A ↗: Passive dashboards that communicate information continuously without requiring human interaction (like the panels in airports). A modality of Layer 1 (Interaction).
SME — subject-matter expert Book III · Appendix A ↗: The human expert whose knowledge is transferred to the agent in the Capability construction scheme. Wingtraining opens with a workshop with the SME and validates the ALPHA phase with them; having an identifiable SME is a criterion of the Capability classification test, and sharing the SME and data model with the containing capacity is the sign that an operation is a feature, not a Capability of its own.
Space Book III · Appendix A ↗: Physical human habitat. The word carries corporeality: office, desk, home, city. In this specification, Space is reserved for humans. The agent’s computational scope is never named Space — it is named Domain (AgencyDomain).
temporality Book III · Appendix A ↗: Regime of the Botlet’s manifestation. A declared attribute, two values: discrete (the Botlet manifests in pulses: it wakes by schedule/trigger/event, acts, rests) and continuous (it manifests sustained: it lives persistent). temporality: continuous ⟺ persistent Layer 3 life (the Botler sustains the execution as long as it lives). “Real time” is not chosen on a delivery channel (push) but by giving the Botlet continuous temporality, which mandates the persistent runtime. A report (snapshot) and a live dashboard are the same manifestation under different temporality → a single runtime.
Tokenization Book III · Appendix A ↗: Replacement of sensitive data with tokens before they reach the cognition model. It lets the agent reason over the data without exposing them to the external cognition provider. A component of the Validation pillar of Trust Infrastructure.
Tool Book III · Appendix A ↗: Instrument that an agent can invoke to touch external systems. Link 9 of the value chain. The contemporary canonical protocol is MCP.
Parallel topology Book III · Appendix A ↗: Canonical model of the relation between the four layers of the Agentive Architecture (Ch. 4). Layers 2 (Cognition) and 3 (Autonomy) are parallel paths between Layer 1 (Interaction) and Layer 4 (Access), not stages in series. An operation that enters through Layer 1 can reach Layer 4 by traversing the Cognition Path (slow, costly, decisive) or the Autonomy Path (fast, cheap, repetitive). The two paths interact (2 ↔︎ 3: cognition delegates to a Botlet, the Botlet escalates fallback to cognition, cognition observes the log) but neither dominates the other. A refoundational model with respect to the linear reading 1 → 2 → 3 → 4.
Trace Book III · Appendix A ↗: End-to-end traceability of an agent operation. It contains identity, capability invoked, tool executed, parameters, result, timestamp, context. A component of the Audit pillar of Trust Infrastructure.
Transparency Book III · Appendix A ↗: Pillar 5 of Trust Infrastructure. The human’s capacity to understand, in real time, what the agent is doing and why — with enough detail to intervene if necessary. It is the pillar that connects the other four: Governance defines what it may do, Audit records what it did, Validation verifies what it is about to do, Resilience ensures it keeps operating; Transparency ensures a human can understand all of the above.
The agent’s three times Book III · Appendix A ↗: Canonical temporal frame of the agent: Preparation (mise en place — refines the catalog, improves capabilities, outside the service window), Attention (interacts with users in real time, the critical path), Engineering (bridge: converts latent capacity into executable capacity for a concrete case). All three are simultaneous but differentiated in regime, urgency, and cognitive economy. The parallel topology describes WHERE each operation lives; the three times describe WHEN the agent operates.
Trust Infrastructure also in Book II Book III · Appendix A ↗: Set of cross-cutting properties that allow an organization to trust that its agents operate with autonomy without losing control. Five pillars: Governance, Audit, Validation, Resilience, Transparency.
Digital twin Book III · Appendix A ↗: A digital twin that reflects in real time the state of a physical system. Canonical pattern for agents to operate over the Carbon World — the agent acts on the twin; the twin propagates to the physical world when it is safe.
ucodex Book III · Appendix A ↗
Validation Book III · Appendix A ↗: Pillar 3 of Trust Infrastructure. The capacity to verify that the agent’s response or action is correct before it affects the world.
Vergis Book III · Appendix A ↗: Proper name of the public reference implementation of AgencyDomains (the platform; the AgencyDomain made operational). Category: Meta-Cognitive Platform. Distributed under AGPL, public repository, AgencyDomains.org. It is the proper name of an instance (same drawer as Soveria, Agentia, ultraPRO), not a type like Botler or Botlet.
Autonomy Path Book III · Appendix A ↗: One of the two paths of the parallel topology (Ch. 4). The path an operation traverses between Layer 1 and Layer 4 passing through Layer 3 (Autonomy) without invoking Layer 2 (Cognition). Its own regime: fast, cheap, repetitive. For the execution of Lets — Botlets over stable patterns and Agentlets over routine judgment —. It is the path that sustains the everyday operation of an AgencyDomain in production and the structural basis of the offline mode when the Botlets are senior.
Cognition Path Book III · Appendix A ↗: One of the two paths of the parallel topology (Ch. 4). The path an operation traverses between Layer 1 and Layer 4 passing through Layer 2 (Cognition) without necessarily passing through Layer 3. Its own regime: slow, costly, decisive. For conversation, new decisions, unanticipated cases, cases where the human needs reasoned interlocution. Economically expensive — the organization uses it sparingly and reserves its capacity for cases where the value justifies it.
Wingtraining Book III · Appendix A ↗: Canonical development scheme of a Capability in five steps: workshop with the SME · creation · customization · ALPHA (validation with the SME) · BETA (validation in real operation). Its development belongs to the delivery practice, not to the canon; the Capability classification test uses it as a criterion because it is the operational proof that a scope is a transfer of knowledge — not an integration or a format.
Wingworking also in Book II Book III · Appendix A ↗: Collaborative practice between human and AI, a generalization of wingcoding (programming with an AI copilot) to any discipline of professional work. It is the methodological frame under which this book was produced — human author and AI agent working in parallel, with an explicit division of roles and a common quality contract. The methodology lives outside the architectural scope of the book; it appears declared in the Colophon as an attribution of the production process.
WorkSpace Book III · Appendix A ↗: The human’s digital work space. The enterprise software industry extended the word Space to WorkSpace (Google Workspace, Microsoft 365, Notion) to name the collection of solutions that digitize what a person does at their physical desk: read email, schedule meetings, write documents, file. WorkSpace is the digital prosthesis of the human Space; both carry the same corporeality of origin and are reserved for humans. The agentive equivalent is Domain (AgencyDomain).
Wrapper / Platform / Core / Infrastructure Book III · Appendix A ↗: The four canonical depths of the AI value chain model. They define, within a given link, how an actor participates:

For agents · this vocabulary as Markdown →The door for agents · llms.txt →The full trilogy · agencydomains.org →

César Obach-Renner · The Agentive World Trilogy