Skip to content
Tutorials Center
My learning

Topics

Lesson 24 of 30 ยท Advanced API Testing

Postman Flows, collaboration and AI features

Tour Postman Flows, workspaces, forks and pull requests, the API Network and the Postbot AI assistant, and learn how to use each one safely and sensibly.

So far you have worked mostly alone, inside one request at a time. Real API work is rarely like that. Teams share collections, several people change the same requests, and some tasks (call this, then that, then decide) are easier to see as a diagram than to read as scripts. Postman has features for each of these: Flows for visual logic, workspaces, forks and pull requests for teamwork, the API Network for finding public APIs, and Postbot, an AI assistant.

This lesson is a guided tour with an honest focus: what each feature is for, when it earns its place, and what to double-check. These are hosted, account-based features. They cannot run on this page, and the product changes often, so names and menu positions below describe the ideas rather than promise exact labels. The runnable examples are stand-ins: a small flow engine and a three-way merge that show how these features work inside.

Five features, one idea

Flows Build API logic as a diagram.
Workspaces Share collections with a team.
Forks and PRs Change safely, review, merge.
API Network Discover public APIs and collections.

Each one moves work from "on my laptop" to "shared and reviewable".

1 / 2

Postman Flows: logic as a diagram

Flows is a visual builder. Instead of writing a chain of scripts, you drag blocks onto a canvas and connect them with lines. A typical flow starts with an input, sends one or more requests, branches on a condition (for example "did the status equal 200?"), reshapes data, and ends in an output. Because every step is a box with a visible connection, a colleague who does not write JavaScript can still read what the flow does.

The block names and their exact settings depend on the Postman version, but the building blocks you will meet are consistent in spirit:

  • Send a request: calls a request from one of your collections (or a new one) and exposes its response.
  • Decide: a condition that sends the flow down a "true" or "false" path.
  • Reshape: picks a field out of a response or builds a new value from several.
  • Repeat or collect: loops over a list, or gathers results from many runs.
  • Output: shows or returns a final value.

Flows suit three jobs. Prototyping a multi-step API scenario before you formalize it. Explaining a workflow to people who do not read code. And small automations where a visual layout is easier to maintain than scripting. For serious regression testing, collections with scripts and Newman remain the dependable, version-controllable choice, covered in Newman, the CLI and CI/CD. Treat Flows as a complement, not a replacement.

How a flow works inside

Under the picture, a flow is a graph: nodes (the blocks) and edges (the lines). To run it, an engine starts at the entry node, does the node's work, and then follows an edge to the next node. A condition node chooses between two edges. That is all the following stand-in does, with a fake API in place of real requests.

JS A tiny flow engine: nodes, edges and a condition

Press Run (or Ctrl + Enter) to see the console output here.

No output.

The two runs take different paths through the same graph, purely because of the condition node. The step limit of 50 is a safeguard: a graph with a loop and no exit would otherwise run forever. Real flow engines have their own protections, but the lesson holds, because a visual flow is still a program and can still have bugs.

Workspaces and collaboration

A workspace is a shared place that holds collections, environments, flows and more. Postman distinguishes kinds of workspace, commonly personal ones, team ones and public ones, with different rules about who can see and edit what. Which kinds you can use depends on your plan and your organization's settings.

Teamwork in a shared workspace means several people see the same collection, so a change by one person reaches the others. That is convenient, and it is also the source of the classic problem: two people edit the same request and one overwrites the other. Postman's answer borrows from software development.

Forks, pull requests and versions

Think of a collection as a shared source file. Postman lets you create a fork: your own copy of a collection, linked to the original. You change your copy freely without disturbing anyone. When you are happy, you open a pull request that proposes your changes to the original. Reviewers see a diff (what changed), comment, and either approve or ask for changes. Once approved, the changes are merged into the original. If you keep your fork up to date, you can also pull in changes others have merged since you forked.

The review loop

Fork Make your own linked copy.
Edit Change it without risk.
Pull request Others review the diff.
Merge Approved changes reach the original.

Nobody breaks the shared collection by accident.

1 / 2

Postman also offers collection versioning, comments and change history so you can see who changed a request and go back. If your team already uses Git, you can additionally keep collections as files in a repository, which is the approach used in Project: automated regression in CI. Choose one source of truth. Keeping a collection edited in two places at once invites exactly the conflicts these features exist to prevent.

How a merge decides: a stand-in

When you merge, the tool compares three versions: the base (the collection when you forked), theirs (the original now) and yours (your fork). A field changed on only one side merges cleanly. A field changed on both sides to different values is a conflict that a person must resolve. The stand-in below applies that rule to the fields of one request.

JS A three-way merge of one request

Press Run (or Ctrl + Enter) to see the console output here.

No output.

The first merge keeps both changes: the new URL and the longer timeout. The second cannot guess whether 3000 or 10000 is right, so it reports a conflict and waits for a human. Real tools handle nested structures, such as scripts and header lists, but the three-way logic is the core.

The API Network

The Postman API Network is a public directory where companies and individuals publish workspaces containing their APIs, collections and documentation. It is a handy place to find a ready-made collection for a service you want to try: you fork it into your own workspace and send requests immediately. Two cautions apply. A published collection is only as trustworthy as its publisher, so read what a request does before running it, especially anything that can create, change or delete data. And never put your own secrets into a collection that lives in a public workspace; use environment variables and keep the current values private, as described in Environments and variables.

Postbot and AI features

Postbot is Postman's AI assistant. Depending on your plan and version, it can suggest tests for a response, explain a response or an error in plain language, help write pre-request or post-response scripts, draft documentation for a request, and answer questions about your collection. You usually reach it from an assistant panel or from a prompt in the script editor.

AI help is useful for speed, but you remain the tester. A few habits keep it safe:

  1. Read every generated test. A test that asserts nothing, or only that the status is 200, passes and proves little. Ask yourself what bug it would catch.
  2. Run it against a failure. Break something on purpose and confirm the generated test actually fails. A test that cannot fail is decoration.
  3. Keep secrets out of prompts. Responses and requests may contain tokens or personal data. Check your organization's rules before sharing real data with any AI service.
  4. Verify claims. An explanation of an error is a hypothesis. Confirm it against the documentation or the server's behavior.
  5. Own the design. The assistant can write assertions quickly, but deciding what matters (the strategy in API testing strategy and test design) is still your job.

Because AI features change quickly and depend on your plan, check what your own workspace offers rather than assuming. Nothing on this page needs AI: every skill in this tutorial works without it.

Choosing what to use

Need Reach for
Try a multi-step scenario visually Flows
Share requests with colleagues A team workspace
Change a shared collection safely A fork and a pull request
Find a ready-made collection for a public API The API Network
Draft tests or explanations quickly Postbot, with review
Automated regression on every commit A collection plus Newman in CI

Check your understanding

Quick check ยท 4 questions

of 4 answered You scored / 4 ยท perfect!

  1. 1. What is a Postman Flow?

    Flows is a visual builder: blocks on a canvas connected by lines, so the logic of a multi-step scenario can be seen at a glance.
  2. 2. Why fork a collection before changing it?

    A fork is your own copy, so you can experiment safely, then ask for review through a pull request before changes reach the original.
  3. 3. In a three-way merge, when is a field reported as a conflict?

    A change on only one side merges cleanly. Two different changes to the same field need a human to decide which to keep.
  4. 4. What is the safest way to use a generated test from an AI assistant?

    A test that cannot fail proves nothing. Review it, and break the API on purpose to see that the test catches the problem.

Summary

  • Flows lets you build multi-step API logic as a diagram of blocks and connections; use it for prototyping, explaining and light automation, and keep collections plus Newman for dependable regression runs.
  • A flow is a graph: an engine follows edges from node to node, and a condition node picks between two paths.
  • Workspaces share collections; forks, pull requests and merges let teams change them safely with review and history.
  • A three-way merge keeps changes made on only one side and reports a conflict when both sides changed the same thing differently.
  • The API Network is a public directory: fork what you need, read before you run, and never publish secrets.
  • Postbot and other AI features speed up writing tests and explanations, but you must review, break-test and own what they produce.
  • The runnable examples are stand-ins (a flow engine and a merge function), not Postman's own implementation.

Next: Project: test suite for a CRUD API.