Skip to content
This repository has been archived by the owner before Nov 9, 2022. It is now read-only.


Switch branches/tags

Name already in use

A tag already exists with the provided branch name. Many Git commands accept both tag and branch names, so creating this branch may cause unexpected behavior. Are you sure you want to create this branch?

Latest commit


Git stats


Failed to load latest commit information.
Latest commit message
Commit time


CircleCI PyPI Code style: black License: MIT

The HTTP Mocking Toolkit (HMT) is a tool that mocks HTTP APIs for use in sandboxes as well as for automated and exploratory testing. It uses a combination of API definitions, recorded traffic and code in order to make crafting mocks as enjoyable as possible.

Chat with us on Gitter to let us know about questions, problems or ideas!

What's in this document


Install via pip (requires Python 3.6+):

pip install hmt

macOS users can install HMT with Homebrew:

brew tap meeshkan/tap
brew install hmt

Debian and Ubuntu users can install HMT with apt:

echo "deb [trusted=yes] all main" | tee -a /etc/apt/sources.list
apt-get -qq update && apt-get install hmt

Getting started with HMT

The basic HMT flow is collect, build and mock.

  1. First, collect data from recorded server traffic and/or OpenAPI specs.
  2. Then, build a schema that unifies these various data sources.
  3. Finally, use this schema to create a mock server of an API.


The quickest way to get an overview of HMT is to complete our interactive tutorial. It walks you through the collect, build, and mock flow - while also covering the concepts necessary for development.

Note: This tutorial has been tested on Python 3.6, 3.7, and 3.8.

After installing HMT, you can begin the tutorial by invoking from the command line:

$ hmt tutorial

Once you've run this, you should see:

    __              __ 
   / /_  ____ ___  / /_
  / __ \/ __ `__ \/ __/
 / / / / / / / / / /_
/_/ /_/_/ /_/ /_/\__/

The tutorial!!
Press ENTER to continue...

If not, it's probably our fault. Please let us know by filing an issue on this repo.

Collect recordings of API traffic

Let's look at how to build a HMT spec. First, you have to collect recordings of server traffic and/or OpenAPI server specs.

To record API traffic, the HMT CLI provides a record mode that captures API traffic using a proxy.

$ hmt record

This command starts HMT as a reverse proxy on the default port of 8000 and creates two directories: logs and specs.

With curl, for example, you can use HMT as a proxy like so:

$ curl http://localhost:8000/

By default, the recording proxy treats the path as the target URL. It then writes a .jsonl file containing logs of all server traffic to the logs directory. All logs are created in the http-types format. This is because HMT's build tool expects all recordings to be represented in a .jsonl file containing recordings represented in the http-types format.

For more information about recording, including direct file writing and kafka streaming, see the recording documentation.

Build a HMT spec from recordings

Using the HMT CLI, you can build an OpenAPI schema from a single .jsonl file, in addition to any existing OpenAPI specs that describe how your service works.

$ hmt build --input-file path/to/recordings.jsonl 

Note: The input file should be in JSON Lines format and every line should be in http-types JSON format. For an example input file, see recordings.jsonl.

Optionally, you can also specify an output directory using the --out flag followed by the path to this directory. By default, HMT will build the new OpenAPI specifications in the specs directory.

Use dash (--input-file -) to read from standard input:

$ hmt build --input-file - < recordings.jsonl

Building modes

You can use a mode flag to indicate how the OpenAPI spec should be built, for example:

hmt build --input-file path/to/recordings.jsonl --mode gen

Supported modes are:

  • gen [default] - infer a schema from the recorded data
  • replay - replay the recorded data based on exact matching

For more information about building, including mixing together the two modes and editing the created OpenAPI schema, see the building documentation.

Mock server traffic using a HMT spec

You can use an OpenAPI spec, such as the one created with hmt build, to create a mock server using HMT.

$ hmt mock path/to/dir/

Note: You can specify a path to the directory your OpenAPI spec is in or a path to one specific file.

For more information about mocking, including adding custom middleware and modifying the mocking schema JIT via an admin API, see the mocking documentation.


Here are some useful tips for building and running HMT from source.

If you run into any issues, please reach out to our team on Gitter.

Getting started

  1. Clone this repository: git clone
  2. Create a virtual environment: python3 -m venv .venv && source .venv/bin/activate
  3. Install dependencies: pip install --upgrade -e '.[dev]'
  4. Install pre-commit hooks to automatically format code as a git hook: pre-commit install


Run all checks:

$ python test


Run tests/ with pytest:

# or
python test

Configuration for pytest is found in pytest.ini.


Formatting is checked by the above mentioned python test command.

To fix formatting:

$ python format


Run style checks:

$ flake8 .


You can run type-checking by installing pyright globally:

$ npm -i -g pyright

And then running:

$ pyright --lib
$ # or
$ python typecheck

Using the Pyright extension is recommended for development in VS Code.

Automated builds

Configuration for CircleCI build pipeline can be found in .circleci/config.yml.

Publishing HMT as a PyPi package

To publish HMT as a PyPi package, complete the following steps:

  1. Bump the version in if the version is the same as in the published package. Commit and push.
  2. Run python test to check that everything works
  3. To build and upload the package, run python upload. Insert PyPI credentials to upload the package to PyPI. The command will also run git tag to tag the commit as a release and push the tags to remote.

To see what the different commands do, see Command classes in


Thanks for your interest in contributing! Please take a look at our development guide for notes on how to develop the package locally. A great way to start contributing is to file an issue or make a pull request.

Code of Conduct

Please note that this project is governed by the Meeshkan Community Code of Conduct. By participating, you agree to abide by its terms.