---
id: "framework-intent-gap-layers"
type: "framework"
source_timestamps: ["00:11:02", "00:16:20"]
tags: ["architecture", "maturity-model"]
related: ["concept-unified-context-infrastructure", "concept-intent-engineering", "concept-machine-readable-okrs", "concept-ai-fluency-vs-activity", "concept-shadow-agents", "action-build-mcp-infrastructure", "action-translate-okrs", "action-hire-workflow-architect"]
steps: ["\"Establish Unified Context Infrastructure (secure", "centralized data access).\"", "\"Deploy a Coherent AI Worker Toolkit (shared", "measurable workflows).\"", "Implement Intent Engineering Proper (machine-readable goals and tradeoff hierarchies)."]
sources: ["s24-prompt-engineering-dead"]
sourceVaultSlug: "s24-prompt-engineering-dead"
originDay: 24
---
# The Three Layers of the Intent Gap

## Overview

A three-layer architectural framework describing the gaps organizations must close to deploy autonomous AI agents at scale. Each layer is necessary; skipping any one prevents the layer above from functioning.

## Layer 1 — Unified Context Infrastructure

Move away from fragmented [[concept-shadow-agents]] and per-team RAG pipelines toward a **composable, vendor-agnostic architecture** (canonically: [[entity-mcp]]) that securely connects AI to organizational data with central governance.

- Concept note: [[concept-unified-context-infrastructure]]
- Action: [[action-build-mcp-infrastructure]]
- Failure mode without it: [[concept-shadow-agents]]

## Layer 2 — Coherent AI Worker Toolkit

Transition from isolated, individual AI tool usage (ChatGPT, Cursor, ad-hoc Claude) to **shared, measurable organizational workflows**.

- Concept note: [[concept-ai-fluency-vs-activity]]
- Failure mode without it: high "activity" with no fluency — the [[claim-copilot-intent-failure|Copilot pattern]].

## Layer 3 — Intent Engineering Proper

The highest layer. Translate human-centric OKRs and implicit cultural values into **explicit, machine-readable parameters, delegation frameworks, and tradeoff hierarchies** that guide autonomous agent decision-making.

- Concept note: [[concept-intent-engineering]]
- Artifact: [[concept-machine-readable-okrs]]
- Action: [[action-translate-okrs]]
- Org owner: [[action-hire-workflow-architect]]
- Failure mode without it: [[claim-klarna-intent-failure|the Klarna pattern]].

## How the Layers Compose

```
┌─────────────────────────────────────────────┐
│  Layer 3: Intent Engineering                │  ← what to want
├─────────────────────────────────────────────┤
│  Layer 2: Coherent AI Worker Toolkit        │  ← how to work
├─────────────────────────────────────────────┤
│  Layer 1: Unified Context Infrastructure    │  ← what to know
└─────────────────────────────────────────────┘
```

A mature enterprise AI deployment moves *up* this stack. Most enterprises in 2025–2026 are stuck somewhere between Layer 1 and Layer 2. Almost none have meaningfully begun Layer 3.



## Related across days
- [[framework-world-model-architectures]]
- [[framework-hybrid-memory-stack]]
- [[framework-2026-builder-practices]]
- [[arc-context-architecture-evolution]]