IDD concept: Specify intent

Intent Shaping

Intent Shaping is the process through which human goals are transformed into intent that is executable, measurable and governable. It begins before implementation and continues throughout delivery, because intent becomes clearer through execution, not only before it.

From goals to governable intent

People begin with goals, not specifications. They describe outcomes, risks, constraints and expectations in natural language and business terms, and those descriptions are rarely precise enough for an autonomous system to act on safely. Intent Shaping turns them into the things a system can work against: success criteria, constraints, boundaries, governance rules and measurable outcomes. The Intent Specification is the artefact; Intent Shaping is the activity that produces it and keeps it accurate.

The work calls for particular skills: recognising unclear intent before it causes misalignment at machine speed, and translating between product thinking, user insight, architecture, security and operations so that the result is coherent. These are among the capabilities that grow in value as AI takes on more of the implementation.

Shaping is continuous

Traditional delivery often treats requirements as something written, approved, implemented and eventually replaced. IDD treats intent as a feedback loop instead. Intent is defined, shaped into something executable, implemented by people or agents, measured against outcomes and constraints, and then refined in light of what was learned.

Each stage of delivery feeds the loop. Implementation reveals missing constraints, unclear success criteria and unexpected edge cases. Measurement through Intent Fidelity reveals where intent and outcome diverge. Governance reviews reveal risks and conflicts. All of it returns to shaping, which is why a specification that has been through several cycles of delivery is usually far better than the first draft, however carefully that draft was written. This ongoing refinement is what IDD calls intent evolution.

Bottom-up discovery, top-down context

Shaping works best when intent moves in two directions at once. From the bottom up, projects discover what actually works, encountering real constraints, real domain rules and real architectural patterns. From the top down, the organisation supplies strategic direction, regulatory constraints, security requirements and shared infrastructure decisions. Bottom-up discovery without top-down context leads to fragmentation; top-down standards without bottom-up feedback lead to architecture disconnected from reality. Intent is shaped where the two meet, and it is there that it can be validated, refined and, where it proves reusable, promoted through the Intent Hierarchy.

Origin of the term

The name intent shaping comes from Calum Gunn, whose contribution is credited in the article that introduced the concept.

Where it comes from

Intent Shaping was introduced in Intent Evolution and Intent Shaping, published on 23 March 2026, and forms Chapter 12 of the ebook.

Frequently asked questions

Is Intent Shaping a one-off activity at the start of a project?No. Some shaping happens before implementation begins, but much of it happens during delivery, as implementation, measurement and governance reviews surface gaps that nobody could have anticipated at the outset.
How is Intent Shaping different from requirements gathering?Requirements gathering usually produces a document that is handed over for implementation. Intent Shaping produces intent that can be executed, measured and governed, and it keeps refining that intent through feedback rather than treating it as finished once approved.
Who is responsible for Intent Shaping?It is shared across roles. Product, UX, architecture, security, engineering and QA each shape the part of the intent they understand best, and the Intent Specification is where their contributions come together.
What is the difference between intent shaping and intent evolution?Intent Shaping is the activity of turning goals into governable intent. Intent evolution describes what happens to that intent over time as shaping continues through cycles of delivery, measurement and review.