What is Mob Elaboration?
Mob Elaboration is a working session from AI-DLC, the AI-driven development lifecycle that AWS published in 2025 (github.com/awslabs/aidlc-workflows). It belongs to the Inception phase — the part of the lifecycle where a business intent is turned into something precise enough to build. The whole cross-functional team sits in one room, an AI agent asks the questions, the team answers, and the agent decomposes the answers into units of work that the team approves one by one.
The name is a deliberate echo of mob programming: everyone works on the same thing, at the same time, in the same space. The difference is what the mob produces. Mob programming produces code. Mob Elaboration produces scope — and, just as importantly, a written record of everything the team could not decide.
Why the AI asks the questions
In classic refinement, a product owner presents a feature and the team pokes at it until the meeting ends. The quality of the outcome depends on who happens to speak up, and the gaps only surface weeks later, in the code. Mob Elaboration inverts the flow. The AI reads the intent and whatever context it has — ideally the repository — and generates the 10–15 questions without which the scope cannot be defined unambiguously: business rules, edge cases, integrations, data, what is out of scope, and how the result will be verified.
Every question carries a default assumption that will apply if nobody answers. That single rule changes the room. An unanswered question is no longer an awkward silence; it is a concrete assumption that will reach the code unless someone objects, with a name and a deadline attached.
The parts that make it work
- One driver, one keyboard. Only one person types to the AI. Everyone else talks to the driver. This keeps the transcript accountable.
- “I don’t know” is a valid answer. The runner challenges answers that sound like guesses, once. A logged “don’t know” with an owner beats a confident guess every time.
- No solution design in the question round. Answers are rules and facts. How to build it comes later, in decomposition.
- The Decisions Board. Every question the team cannot answer within its mandate is logged with a temporary assumption, an owner, and a deadline. The count of unowned items is read out loud at the close — it is the headline of the session.
From answers to units of work
After a break, the AI proposes units of work, one at a time. Each has a goal, scope, out-of-scope list, testable acceptance criteria, dependencies, and the temporary assumptions it relies on. Each is labelled: autonomous, meaning an AI agent with repository access could build and test it in one working session, or needs-human, meaning it is blocked on an open decision. The team accepts, amends, or rejects each unit. Unit boundaries are an architecture decision, so nothing is labelled autonomous until an architect or tech lead signs off that the boundaries match how the code is actually coupled.
How it differs from classic refinement
| Classic refinement | Mob Elaboration |
|---|---|
| Team questions the product owner, ad hoc | AI generates the question list systematically |
| Gaps surface during development | Gaps become logged assumptions with owners, in the room |
| Output: estimated backlog items | Output: units an agent can start building, plus artifacts |
| Unanswered questions evaporate | Every one goes on the Decisions Board |
| Success is a feeling | A metrics log, filled in again after Construction |
What it is not
It is not a discovery workshop — without a written intent the session should be postponed, not improvised. It is not a design session — the AI parks solution talk until decomposition. And the AI is not an oracle: it runs the process and keeps the memory, but the product decisions stay with the humans, on the record.
The session runs about two and a half hours with hard stops, and ends with six markdown artifacts saved to the repository and a metrics table. Whether all this actually shortens the path from intent to pull request is an open question — which is exactly why this site collects session metrics from pilot teams.