1. Docs
  2. Resources
  3. Deploys & Repos

Deploys & Repos

Deploys & Repos answers two questions per service and environment: which version is running, and did a pipeline test it after it was deployed? Your CI tells SmartestQA about each deploy; SmartestQA matches it with the pipeline runs that followed.

6 min read Updated CustomersAdmins

In short

On this page you can

  • See the version (commit and tag) each service runs in DEV, TEST, PRP and your own environments.
  • Spot deploys that no pipeline tested, shown as Not tested.
  • Open a deploy to read its commits, changed files and the pipeline runs after it.
  • Connect GitHub, GitLab, Bitbucket, Azure DevOps or Gitea repositories, with or without a token.

How to use it

  1. Open Repositories and click Add repository.
  2. Copy the deploy token and the CI step from CI code & deploy token.
  3. Add the step after your deploy step in CI and run one deploy.
  4. Run a pipeline in the same environment, then check Currently running versions.

At a glance

Where
Main menu › Deploys & Repos. Tabs: Overview and Repositories.
Who
Admins and Users with access to the project. Admins also see the repositories under Integrations.
You need
A Git repository URL and a CI system that can send an HTTP request after a deploy. A read-only token is optional.
Time
About 10 minutes: add the repository, store the deploy token in CI, add one step.
Produces
Deploy records per service and environment, with the previous version, commits and the pipeline runs that tested it.
Used by
The KPI cards on this page and Autonomous Agents that start when a deploy succeeds.
Project types
The same for WEB, MOBILE and DESKTOP projects. A deploy is matched with pipeline runs by environment, not by project type.

Start from your question#

Where Deploys & Repos fits#

The main menu is the order of work. Deploys & Repos sits beside it: it does not create tests. It watches what reaches each environment and checks whether the pipelines you built in Schedule ran against it afterwards.

  1. 1DiscoverPages, controls, rules
  2. 2AnalyzeRequirements
  3. 3AutomateTest cases
  4. 4SchedulePipelines per environment
  5. 5ReportResults of each run
  6. 6MaintenanceHeal and review

A deploy record comes from your CI. The “tested after” state comes from pipeline runs in Schedule. A click on a run opens it in Report. A failed run can become a task in Tasks.

Why use it#

  • You see untested releases. A cell turns amber with Not tested when a version reached an environment and no pipeline ran there since.
  • You know what changed. With a token, each deploy lists the commits and changed files since the previous successful deploy in the same environment.
  • Several services, one view. A project can have many repositories (web, api, auth…). Each one is a row; each environment is a column.
  • Agents can react to deploys. Every successful deploy is passed to Autonomous Agents, so an agent can run checks right after a release.
  • It never blocks your release. The CI step ends with || true: if SmartestQA cannot be reached, the deploy still succeeds.

The building blocks#

Repository
A Git repository connected to the project, for example https://github.com/acme/web-frontend. It has a provider, a service name, an optional access token and an environment mapping.
Service name
The short key deploys are shown under, such as web or api. Unique in the project.
Environment
Where a version runs: DEV, TEST, PRP, or a name you add such as STAGING. Pipeline runs carry DEV, TEST or PRP.
Environment mapping
Which environments you expect this repository in, with an optional branch and start URL for reference. It decides which cells show Not mapped.
Deploy token
A secret that starts with sdt_. It lets your CI record deploys for one repository and nothing else. More.
Access token
An optional read-only token for your Git provider. With it, SmartestQA reads commits and changed files between two versions. More.
Deploy
One record: repository, environment, commit SHA, status (Succeeded, Failed, In progress, Cancelled), version, branch, who deployed it and a link to the CI run.
Tested after
The pipeline runs in the same environment that started after a successful deploy and before the next one. Formula.

The screen#

Deploys & Repos › Overview with the four KPI cards, two repositories (web, api) and DEV, TEST, PRP columns, one amber Not tested cell
Deploys & Repos › Overview.
AreaWhat it is forGo there when
Currently running versionsThe last successful deploy per service and environment, with its test state.Before a release decision, or to see what a tester is testing.
Deploy historyEvery deploy record, newest first, filterable by environment, service and period.You look for a failed or cancelled deploy, or for who deployed what.
Deploy detailPrevious version, commits, changed files and the pipeline runs after the deploy.A run went red and you want to know what changed.
RepositoriesConnections, environment mapping, deploy token and CI code.You set up a new service or rotate a token.

No project selected? The page shows “Select a project from the header to manage its agents.” Pick a project in the header.

Before you start#

You needWhyCheck it
The repository URLTo create the connection and build commit links.It opens in your browser and contains owner and repository.
Access to your CI settingsTo store the deploy token as a secret and add one step.You can add a secret or credential to the pipeline.
Pipelines in Schedule that run per environmentOnly pipeline runs count as “tested after”.Your pipeline’s environment is DEV, TEST or PRP.
Optional: a read-only Git tokenTo see commits and changed files.See the token types in Connection modes.

Best practices#

Do

  • Use one repository per deployable service, with short service names (web, api).
  • Send the same environment names your pipelines use: DEV, TEST, PRP.
  • Send the full commit SHA and, if you tag releases, the tag as version.
  • Run a pipeline in that environment after every deploy, for example a smoke pipeline.
  • Use a fine-grained, read-only token when you connect with a token.

Don’t

  • Paste the deploy token into CI code. Store it as a secret.
  • Give the access token write rights. Reading is enough.
  • Report deploys to environment names no pipeline uses, unless you only want the version record.
  • Reuse one repository record for two services: the deploys would mix in one row.

In this section#

Didn’t find your answer?Write to support with your project name and what you see on screen.

info@smartestqa.ai