Structured Output Contracts

Structured output contracts define which machine-readable outputs are meant to be stable enough for automation.

Output Contract Model

flowchart LR
    Command[CLI or API command] --> Json[Structured output]
    Json --> Automation[Automation and CI]

This output-contract diagram shows why structured output deserves its own contract page. It is the surface automation should parse when Atlas explicitly documents it as stable.

Stability Logic

flowchart TD
    Stable[Documented structured output] --> Parse[Safe to parse]
    Unstable[Undocumented text output] --> Human[Human-only interpretation]

This stability logic keeps the parsing rule simple: documented structured output is the contract, while human-readable text remains descriptive.

Main Promise

If Atlas documents a structured output surface and tests it, automation should prefer that surface over screen-scraped human text.

Stability

Only structured outputs that Atlas documents as contracts should be treated as stable automation inputs. Human-readable text remains descriptive and may change without compatibility guarantees.

Purpose

This page defines the Atlas contract expectations for structured output contracts. Use it when you need the explicit compatibility promise rather than a workflow narrative.