> ## Documentation Index
> Fetch the complete documentation index at: https://lab.pollack.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Knowledge-Directed Execution

> The thesis: knowledge + structured execution > model

## The Thesis

> **Knowledge + structured execution > model**

Agent reliability improves more from giving agents the right knowledge and constraining their execution than from switching to a larger model.

## What This Means

### Knowledge

Not "more data" — **curated, structured domain knowledge** delivered to the agent at the right time:

* Which testing patterns work for this framework
* What dependencies are available and how to use them
* What the project conventions are
* What common failure modes look like

### Structured Execution

Not "better prompts" — **infrastructure that shapes agent behavior**:

* Deterministic preprocessing before the LLM acts
* Tool configuration that guides tool selection
* Execution loops with built-in checkpoints
* Judge feedback that catches failures early

### > Model

This doesn't mean models don't matter. It means that for a given model, you get more reliability improvement from knowledge and execution infrastructure than from upgrading to the next model tier.

## Evidence

### Code Coverage v1

The [first experiment](/experiments/code-coverage-v1) showed two independent axes of improvement — knowledge injection and prompt hardening — both of which operate on infrastructure, not model choice. The PetClinic "model floor" (92-94% coverage regardless of variant) demonstrates that the model's prior knowledge creates a ceiling that only infrastructure can differentiate.

### SkillsBench (External)

[SkillsBench](https://arxiv.org/abs/2602.12670) (Feb 2026) found that 2-3 curated skills improve agent performance by +16.2 percentage points on average. Comprehensive skills actually *decrease* performance by -2.9pp. This validates "curated > comprehensive" — structure matters.

### Stripe Convergence

Stripe's [Minions paper](https://arxiv.org/abs/2402.15678) independently arrived at similar conclusions: "the walls matter more than the model." Their multi-agent system improves reliability through structured task decomposition, not model upgrades.

## The Equation

```
Agent Reliability = f(Knowledge Quality × Execution Structure × Model Capability)
```

Current industry focus is almost entirely on Model Capability. This lab focuses on the first two terms, where the marginal returns are higher.

## Naming History

This concept has gone through several names:

| Name                                      | Status                            |
| ----------------------------------------- | --------------------------------- |
| "Infrastructure over prompts"             | Early framing, too narrow         |
| "Knowledge-directed execution"            | Current, captures both components |
| "Curated opinions + structured execution" | Verbose but precise               |
| "The walls matter more than the model"    | Stripe's phrasing, resonant       |

See [journal/2026-03-02-naming-the-thesis.md](https://github.com/markpollack) for the full naming discussion.

## How to Apply

If you're building agent systems:

1. **Start with knowledge** — What does your agent need to know that it doesn't?
2. **Structure the delivery** — Skills > flat files > nothing
3. **Add execution constraints** — Deterministic preprocessing, judge feedback loops
4. **Then consider the model** — Upgrade only after infrastructure is solid