What You Need to Know
A long, productive conversation is not a deliverable. Its findings, decisions and usable text sit among dead ends and superseded drafts. Two deliberate acts turn it into one, and neither happens on its own. Curation extracts what the deliverable needs. Format selection gives it the container its reader will use: an inline chat response, a structured table or an artifact. An artifact is substantial, standalone content that Claude puts in a dedicated window, separate from the conversation.
A plumbing firm's office manager has spent weeks planning winter work with Claude. Three outputs are due. A customer wants to know today whether the boiler service includes a flue check. The owner must choose between three boiler suppliers on price, warranty and delivery time. The plumbers need the winter call-out procedure, which the firm revises every autumn.
From Working Material to Work Product
Curation starts from the deliverable, not the conversation: who reads it, what they do with it, and whether it will be reused or revised. Then extract exactly what that reader needs and shape it into its final container. Forwarding the whole conversation, or leaving the deliverable scattered across it, delegates the curation to the reader. The reader won't do it, so next month's colleague scrolls thirty screenfuls hunting for decisions.
Most of a working conversation is scaffolding, so curation is selective by design. The scaffolding got you to the decisions, and the deliverable owes its reader the decisions alone. The plumbers get the final call-out steps, not the drafts.
The Three Format Homes
The Help Center's definition of artifacts doubles as the selection rule. Artifacts suit significant, self-contained content you are likely to edit, iterate on or reuse outside the conversation. Invert it and the other two homes define themselves.
An inline chat response fits quick, one-off answers consumed in the moment: a fact, a short explanation, text for immediate paste-and-send. A structured table fits anything the reader compares or scans: options against criteria, items with consistent attributes, schedules. If the reader's eye wants rows and columns, prose is the wrong container. An artifact fits a standalone document, plan or reference page with its own life, revised, reused or shared as a thing in itself.
The choice follows the reader's use: ask-and-move-on wants inline, compare-and-scan wants a table, reuse-and-revise wants an artifact. So the flue-check answer goes inline, the supplier choice becomes a table with the criteria as columns, and the call-out procedure becomes an artifact.
The Artifact's Life
The artifact earns its place through what happens after the first draft. You can ask Claude to change it, or edit a document, deck or design directly. The deliverable then evolves as one thing, not as fragments re-pasted through the chat. To try a different direction without losing what you have, edit an earlier message. That creates a different version of the chat with its own artifacts, and the original stays where it was. Sharing gives the artifact a link others can open.
For shared work, a colleague picking up a runbook next year opens the current document. Next autumn, the plumbing firm's manager opens one revised call-out procedure. The format choice is also reversible and always in your hands. If useful content arrived inline, asking Claude to turn it into an artifact takes one message.
What Each Mismatch Costs
Each wrong format has a predictable cost. A reusable runbook left as loose chat prose gets lost and re-created. A comparison delivered as paragraphs makes every reader rebuild the table in their head, and some rebuild it wrongly. Everything in a table flattens narrative judgement into cells that cannot hold it. Everything in an artifact adds ceremony to throwaway answers, such as the flue-check reply. Format follows the reader's use, never habit.
In the exam
The exam tests format selection through these failure modes. This objective chooses the shape of the output. Choosing the product feature or model for the task is Domain 3: a Project for recurring context, Research for a cited report, which model to use. A format item may mention those features in its scenery. The keyed judgement is always about the container the reader receives.
Key Concept
For every output, ask one question: what does the reader do with this? Read it once, compare across it, or come back to it? The answer is the format, and it beats every default. Exam items reward matching each output to its use and penalise both lazy prose and reflexive over-structuring.
Current state: artifacts changed in September 2026
Anthropic reworked artifacts on 16 September 2026. They are now designs, decks, documents, dashboards and small tools made in any conversation. Those made in a chat before that date are "legacy": they can still be opened and shared, but no more can be made. The earlier Help Center article described a version selector for switching versions. The current article, checked 28 September 2026, no longer documents one. The exam guide names artifacts only as an output format, so no keyed answer turns on that control. Where these lessons talk about comparing versions, they mean the branching described above.
Exam Traps
Keying a format question on Projects, Research mode or model choice
Feature and model selection for the task is Domain 3 territory. When a stem asks how information should be organised or delivered, the keyed answer is about the output's container: inline, table or artifact, matched to the reader's use.
Defaulting every output to the same container
Everything-as-an-artifact adds ceremony to throwaway answers, and everything-as-a-table flattens narrative judgement into cells. Format-by-habit options sound systematic and safe, which is exactly what makes them tempting distractors; the keyed answer varies the format with the use.
Sharing the whole conversation instead of curating the deliverable
Forwarding weeks of working material delegates the curation to readers who will not do it: the decisions stay buried among dead ends and superseded drafts. The keyed answer extracts what each reader needs into the container that fits their use.
Practice Scenario
A swimming club's gala organiser has weeks of Claude conversations about the annual gala. Three outputs are now needed: the agreed start time for tonight's committee question, a comparison of three candidate dates against pool availability, volunteer numbers and fixture clashes, and the gala-day runbook that will be revised and reused every year. She wants each in the simplest form its use allows, and her successor must be able to pick up the runbook and edit it. Which formats fit?
Build Exercise
Curate One Deliverable Out of a Working Chat
What you'll learn
- How to identify what a conversation owes its readers
- How to match each output to inline, table or artifact by use
- How to iterate a deliverable as one evolving artifact
- How to test a format choice by retrieving the output later
- Pick a real multi-session piece of work you have discussed with Claude: an event, a project, a recurring report
Why: Curation only shows its value against a conversation long enough to have buried things in.
You should see: A working conversation with findings, decisions and drafts scattered through it.
- List the three outputs it actually owes: something consumed in the moment, something a reader will compare, and something that will be reused or revised
Why: Starting from the deliverables, not the conversation, is the curation habit itself.
You should see: Three one-line output descriptions, each named with its reader and their use.
- Get the quick answer inline, and direct the comparison into a table with the decision criteria as columns
Why: Practising the two lighter formats shows how little ceremony the in-the-moment and compare-and-scan outputs need.
You should see: A one-line inline answer, and a table the reader can scan without rebuilding it mentally.
- Have the reusable piece created as an artifact, ask for one revision, then edit your earlier message to branch the chat and confirm the first draft is still there
Why: The artifact earns its place through iteration; a revision you can compare with the first draft proves the deliverable now evolves in place.
You should see: One artifact revised in place, with the first draft still reachable on its own branch, not fragments re-pasted through the chat.
- A week later, retrieve all three outputs and note which format made retrieval trivial
Why: Retrieval is the honest test of a format choice; the cost of a wrong home only shows when you go back for the content.
You should see: A short note: what was instant, what needed hunting, and what that says about the format chosen.
Sources
- What are artifacts and how do I use them? — Anthropic
- Share artifacts — Anthropic
- Claude Certified Associate – Foundations Exam Guide — Anthropic