LAYER 01 · Build the mental model

What is a harness, really?

The model proposes the next step; the harness turns it into controlled, recoverable action

12 min2 source anchorsupstream@47f9438
model + context + tools + state + control + surfaces
The question this chapter resolves

Why can the same model feel like a different system inside another agent product?

Build intuition

Think of the model as a brilliant new pilot. The harness is not one cockpit button; it is the aircraft, instruments, checklists, air traffic control, flight recorder, and maintenance regime. The pilot chooses; the system makes the choice executable and accountable.

MECHANISM

Mechanism

DeepSeek Harness assigns the complexity around model calls to plugin-owned capabilities: sessions preserve facts, prompt assembly builds requests, tools execute actions, policy and sandboxes constrain them, and client surfaces present results.

The design goal is not to make the model intrinsically smarter. It is to keep models replaceable, runs replayable, capabilities composable, and failures convergent.

Step 1 of 4

Receive intent

Web, CLI, ACP, or SDK delivers the same agent input.

INVARIANTS

What extensions must preserve

  • Every model-visible fact must be reconstructable from the log
  • External action must cross a harness-owned capability boundary
  • Replaceable capabilities must not leak concrete providers into the agent loop
FAILURE MODES

The tempting wrong turns

  • ×Equating tool calling with a production agent
  • ×Stuffing every state into the messages array
  • ×Using prompt wording as a substitute for real permissions and sandboxing
VERIFY IN SOURCE

Do not trust the conclusion—verify it

These anchors are pinned to official deepseek-harness@47f9438; account for later breaking changes when reading current upstream.

KNOWLEDGE CHECK

Pause for ten seconds before revealing

Where is the responsibility boundary between model and harness?

Why the next chapter follows

Next we examine why DeepSeek Harness organizes these responsibilities as plugins instead of growing the central loop.

Learn DeepSeek Harness

An independent learning project. Explanations derive from source; upstream remains authoritative.