# The system around the model deserves its own tests

A model does not work alone. The surrounding system decides which tools it can use, where it keeps state, how it hands work to another agent and what counts as finished.

ID: research-testing-the-system
Author: Dominic Grieco
Published: 2026-09-07T17:56:36.430Z
Updated: 2026-09-07T17:56:36.430Z
Canonical: https://dom.sitesagesolutions.com/research/testing-the-system-around-the-model

Natural-Language Agent Harnesses asks whether that system's policy can become an explicit document rather than stay buried in controller code. The paper pairs those documents with a runtime that turns them into agent calls, handoffs, checks and saved outputs.

## Paper
Natural-Language Agent Harnesses, v2

## Paper authors
Linyue Pan, Lexiao Zou, Shuo Guo, Jingchen Ni, Hai-Tao Zheng

## Paper URL
https://arxiv.org/html/2603.25723v2

## Read at
2026-09-06T19:32:32.525715+00:00

## Paper claims
The authors report comparable task outcomes across several benchmark families while making the policy easier to inspect.

## Interpretation
The distinction I find useful is between policy and mechanism. A document can describe how work should proceed. Code still needs to enforce exact behavior such as parsing, tool execution and isolation. Writing a clear rule is not proof that the system follows it.

For our own work, this suggests a useful test: can we show where a rule affects the run, and can we detect when it fails? That seems more valuable than judging an agent workflow by how thorough its instructions look.

## Limits
Natural-language ambiguity remains a risk. Exact mechanisms need code enforcement.

## Sources
- Natural-Language Agent Harnesses, v2: https://arxiv.org/html/2603.25723v2