Proton Docs
Proton is Atomic's platform for test and process automation. In Proton, you register your automations, set up the data for each one, run them on your company's machines or in the cloud, follow the results step by step, and measure the health of what is automated, with help from AI.
This guide shows how to use the Proton 5 portal. For the automation code, see the Proton Cloud SDK; to prepare the machines, see Environment Setup; to take what happens in the automations to the company's monitoring and data warehouse, see Integrations.
What you can do
- Build automations from reusable components, with the data for each run kept in datasets.
- Run now, schedule, or create recurring runs, on a runner or a cluster, with the environment (staging, production) chosen when you trigger the run.
- Follow each run live, with the step timeline, the log, the evidence, and the AI failure diagnosis.
- Measure quality and operations: success rate, most frequent failures, schedule, log search, and the AI analyses.
- Connect to the business: processes, traceability, automation documentation, and the integration with Jira, Xray, and Zephyr.
- Operate robots with the exception queue, when the organization uses Proton for RPA.
Key concepts
| Concept | What it is |
|---|---|
| Organization | The company in Proton. Each organization has its own data, users, profiles, and plan. A person can belong to more than one. |
| System | What is automated: a portal, an API, SAP. Each system is linked to the git repository that holds its code. |
| Git repository | Where the automation code lives, on GitHub, GitLab, or Bitbucket. Proton reads the code through the registered git integration. |
| Component | A reusable automation step, such as "Log in" or "Create order". The component name is the same in Proton and in the code. |
| Parameter | A piece of data that the component receives (input, in_) or returns (output, out_). |
| Automation | A sequence of components that tests or performs a process, such as "Sales order". |
| Dataset | A version of the automation with the steps in order and the value of each parameter. An automation can have several datasets, one for each set of data. |
| Environment | The set of variables for a place where the automation runs (addresses, users, passwords), such as Staging or Production. |
| Automation group | The datasets that run together, in the Automation Center. A group is either for testing or for operations. |
| Cycle | A round of runs of a group, for example a sprint or a release of the system. |
| Run | One execution of a dataset on a runner, with the status, the log, the evidence, and the result of each step. |
| Runner | The machine that runs the automations, with Proton Runner installed. |
| Cluster | A set of runners that share the runs among themselves. |
The path of an automation, end to end:
- The component code lives in the project's git repository, linked to a system.
- In Proton, the components are registered with their parameters, and the automation puts the components together in order.
- Each dataset of the automation holds the parameter values.
- The dataset is added to an automation group and run on a runner.
- The runner downloads the code, runs the steps, and sends back the result, which shows up in Proton.
Quality and Operation
Proton serves two uses, which Atomic enables per organization:
- Quality (QA): test automation, with cycles, run validation, and integration with test management.
- Operation (RPA): robots that run business processes, with the exception queue for handling a robot that failed or didn't run.
With both enabled, each automation group states what it is for (testing or operations), and the metrics screens get the lens "All", "Quality" or "Operation".
Plan, permissions, and integrations
Some features depend on the organization's plan (the Tier, shown at the bottom of the menu), on the person's profile, or on a registered integration, such as Jira. Without the plan, the screen shows that the feature isn't enabled; without the permission, it tells you that you don't have access. The pages in this guide say when something depends on this.
The portal menu
| Menu group | What it contains |
|---|---|
| Home | The dashboard with the numbers for the period, the AI insights, and shortcuts. |
| Insights | AI analyses, Execution Analytics, Log search, Automation Assets, and Reports. |
| Automations | Automation Center, Schedule, Automations, Component, General Parameter, Environments, and Tags. See Automation Center, Schedule, and Automations. |
| Operation | Robot Exceptions (with the RPA module). |
| Processes | Process map, Traceability, People and Business areas. |
| Resources | Files, Data Tables, and Mobile Apps. |
| Access & Teams | Username, Profile, and Team. |
| Runners | Manage and Cluster. See Runners and clusters. |
| Settings | System, Settings, Git Repository, and Integrations. See Systems and repositories and Integrations. |
| Financial | Licenses and AI usage. |
| Help | What's new, Proton Docs, Proton Runner, and Proton Support. Support tickets are handled in MySignal: see Proton Support. |
Where to start
- Access Proton and learn how to navigate.
- Follow your first automation, end to end.
- For the machines that run the automations, see Environment Setup.
- To use Proton from Claude or ChatGPT, see Proton MCP.
Proton terms are explained in the Glossary.