EXHIBIT 07 / AGENT SYSTEMS

Who keeps an agent’s loop turning—and decides when it stops?

The harness is the turn-by-turn program around a model call. It assembles context, invokes the model, validates tool requests, handles results, persists messages, and decides whether to continue, pause, or stop.

THE SHORT ANSWERThe model supplies outputs. The harness owns the run.

5–8 MIN · MANUAL CONTROLS · NO ACCOUNT
ROOM 01

One loop has jobs the model cannot perform by writing text.

  1. 01 / ASSEMBLEBuild this turn’s input

    The harness chooses which instructions, messages, memory, tool descriptions, and results the model receives.

  2. 02 / INTERPRETClassify the output

    A response may be user-facing text, a tool request, invalid structure, or a signal that another software rule must handle.

  3. 03 / ORCHESTRATERun controlled work

    The harness connects validation, approvals, execution, result delivery, streaming, retries, and cancellation.

  4. 04 / PERSISTMake the run recoverable

    Messages and execution state can be stored so software—not the model’s hidden state—can resume after a disconnect.

IMPORTANT BOUNDARY

“Harness” names a mechanism layer, not one required product. Some systems combine it with runtime, UI, memory, and scheduling; others expose a smaller loop that callers assemble themselves.

ROOM 02

Follow the control hand-offs in a single tool turn.

This record isolates ownership. It is not a transcript from Cloudflare or any model provider, and nothing advances automatically.

EDITORIAL TEACHING TRACENOT A LIVE MODEL OR SERVER LOG

HARNESS

Assemble the envelope

The program selects the system instruction, recent messages, relevant memory, and only the tools enabled for this run.

WHAT THE RECORD MAY SAY

The model sees an input assembled by software.

Step 1 / 4

Keyboard: focus the trace and use ← or →. Nothing advances automatically.

ROOM 03

Operate the three gates in front of a tool.

The browser checks the gates in order and stops at the first failure. This is a deterministic rule demonstration, not a security product.

LIVE BROWSER RULE EVALUATIONNO EXTERNAL EFFECT

SCENARIOCandidate request: sendInvoice({ customerId: “C-104”, amount: 850 }).

DECISION / 2 OF 3 GATES ON

The action must not proceed.

Reject: executable code must not receive unchecked arguments.

A production harness also needs authentication, authorization scope, durable idempotency, failure recovery, observability, and a trustworthy tool implementation.

ROOM 04

The source defines the layer; the museum exposes its consequences.

DOCUMENTED

The harness owns turn-by-turn work

Cloudflare’s documentation lists prompt construction, model execution, tool orchestration, message persistence, streaming, recovery, and continuation among harness responsibilities.

BROWSER-COMPUTED

The three-gate verdict is real

When you change a switch, this page really recomputes the first failing rule and updates the result. It does not execute the displayed tool.

NOT CLAIMED

Complete production safety

Three switches cannot establish correct authorization, safe retries, reliable persistence, or the behavior of any named agent product.

OFFICIAL & PRIMARY SOURCES

See exactly what the museum reviewed.

Agent harnesses Cloudflare Docs · key-sections · checked 2026-09-07 The definition of an agent harness and its listed responsibilities; the rest of this page and the wider Agents documentation were not read. Function calling and other API updates OpenAI · unrecorded
NEXT IN THE AGENT SYSTEMS WINGSee what MCP standardizes—and what it leaves to the host →