We’ve got a lot planned!
This roadmap is directly connected to our project management software. Our roadmap is serious, not just for show!
We will provide build connectors so that projects using technologies like docker-compose, Google Cloud Platform and next.js can benefit from one-click testing.
Allow APIs to authenticate for tests and validate correct 'unauthorized' responses.
Determination of fields that cause errors
The idea is to find out if the presence of a field in a request, or a particular value, or a particular range of values always cause an error.
Adaptive query selection
If a particular query (query, mutation, or subscription) returns only errors after N iterations, stop calling it. Try to process if the field no longer exists or why it won’t respond (if it needs an argument or auth). Exclude it from test reports.
Adaptive field value generation
Improve data generated for fields like DateTime for more granular correctness. i.e. Dates, Times, DateTimes, ISO DateTimes, Native DateTimes, etc.
Improve Field classifier
Generating realistic fake data from APIs. The more knowledge about data type we have the better data we can generate.
GraphQL is pivotal in the world of APIs. Launch first class testing support for GraphQL.
Shipping a second testing algorithm that runs 10x as many tests, runs stateful test cases, grouping of similar errors, and provides human interpretation of the underlying issue casing the bug.
Plugin API for enhanced automated testing
Meeshkan will expose a rich set of parameters to guide the automated testing process.
Automated bug fixes
Meeshkan will propose pull requests to fix certain defects it identifies through automated testing.
Extract data from unstructured responses and logs
Sometimes logs and responses have clues on what exactly caused errors.
Automatically prioritizing bug severity
Expand source-control offerings
Allow developers to import projects from Gitlab, Bitbucket, and all major public clouds
Develop the Meeshkan Recorder.
First-class mocking of dependecies
Release a marketplace of mocks to sub into your project for a bullet-proof testing environement.
This will provide feedback once a bug that was spotted previously, is fixed and create a test case to prevent regressions!
We will provide full support for testing bi-directional gRPC APIs
If we define a reward model, we may use reinforcement learning techniques to generate better data. Actions could be changing or preserving of field values.
Processing error messages
Error messages are a valuable source of information for shrinking and grouping.
Field value collection and generation
Get statistics on field values with particular names or types.