Skip to content
Skip to content
PromptingMaturity: Emerging

Semantic Anchors & Contracts

Use one known term to steer the model. Define your own meaning when the project needs it.

1 min read

What it is

Semantic Anchors are public terms you can say to an LLM and expect a useful result. The term pulls in a known pattern. For example, arc42 points to an architecture template, BLUF means lead with the answer, and Gherkin points to acceptance tests.

What contracts add

Contracts are local. They tell the team what a term means in this repository. That keeps the anchor useful and keeps the output consistent.

Semantic cluster

Hover a node

Semantic anchor

One term, one cluster

arc42

Architecture template

Semantic contract

Write the local meaning down in AGENTS.md or CLAUDE.md.

A public term opens the cluster. A contract fixes the meaning in your repo.

Anchor → Contract in two steps

Use a known term first, then fix its meaning for your project.

Step 1/2

Anchor: invoke a known term

Use an established term so the model activates its full knowledge.

String prompt = "Use TDD, London School for the transfer service.";

The model applies outside-in design, mocks at boundaries, no extra explanation needed.

Sources

Related reading