Test Automation Platform

Cross-team quality intelligence platform for automated testing. It ingests test results, normalizes execution history, computes quality signals, manages test cases and requirements, classifies failures with AI, and runs agentic workflows for impact analysis, test generation, and flaky-test healing.

Demo

What's in v2.0

Commit b4640398780f0f339bb3a9bf5e83b271ab4e1b01 initializes the v2.0 platform revamp:

platform-agent: Agent Hub for workflow orchestration, node routing, PR impact analysis, test generation, automation generation, review decisions, and flaky-test fixes.
platform-storage: shared blob storage with filesystem, MinIO, and S3-compatible backends.
platform-integration-config: integration adapter registry with initial Jira Cloud and GitHub adapter skeletons.
Canonical quality model: requirements, test cases, traceability edges, test suites, manual test runs, agent workflows, review requests, token budgets, GitHub PR tracking, and impact analyses.
Portal expansion: requirements, PR analyses, impact analyses, project settings, test case management, test runs, flaky tests, API keys, and AI settings.
Database migrations extended through V44.

Architecture

                          Test Automation Platform

  Test frameworks and CI
  JUnit 5 / TestNG / Cucumber / Allure / Playwright / Newman / k6 / Gatling / JMeter
                 |
                 v
  +----------------------+       Kafka        +----------------------+
  | platform-ingestion   |  test.results.raw  | platform-analytics   |
  | :8081                +------------------->| :8082                |
  | parse, normalize, DB |                    | flakiness, TIA, logs |
  +----------+-----------+                    +----------+-----------+
             |                                           |
             |                                           v
             |                                +----------------------+
             |                                | platform-ai          |
             |                                | :8084                |
             |                                | Claude/OpenAI        |
             |                                +----------------------+
             |
             v
  +----------------------+                    +----------------------+
  | PostgreSQL 17        |<------------------>| platform-agent       |
  | source of truth      |                    | :8086, host :8087   |
  +----------------------+                    | agent workflows      |
             ^                                +----------+-----------+
             |                                           |
  +----------+-----------+                    +----------v-----------+
  | platform-portal      |                    | MinIO / S3 / FS      |
  | :8085                |                    | artifacts, diffs,    |
  | BFF + React SPA      |                    | checkpoints          |
  +----------------------+                    +----------------------+

  Supporting services: Redis, OpenSearch, Logstash, Loki, Promtail, Grafana,
  Prometheus, and Flyway.

Maven modules

Module	Port	Role
`platform-common`	-	Shared DTOs, enums, Kafka topics, agent contracts, integration contracts, storage contracts
`platform-storage`	-	BlobStore auto-configuration for filesystem, MinIO, and S3-compatible storage
`platform-core`	-	JPA domain, repositories, Flyway migrations `V1` through `V44`
`platform-ingestion`	8081	Result ingestion, parsers, project/team management, requirements, integrations, TCM APIs
`platform-analytics`	8082	Flakiness, trends, quality gates, release reports, Test Impact Analysis, alerting, log search
`platform-integration`	8083	Jira ticket lifecycle for repeated failures and flaky tests
`platform-integration-config`	-	Project-scoped integration adapter registry and adapter contracts
`platform-ai`	8084	Failure classification, batch analysis, AI settings (all models via the LiteLLM gateway)
`platform-portal`	8085	Spring Boot BFF plus embedded React SPA
`platform-agent`	8086	Agent Hub, node registry, workflows, webhooks, review gateway, impact/test-generation/healing flows
`platform-testkit`	-	Rich Java instrumentation for JUnit 5, TestNG, Cucumber, Playwright-assisted tests
`platform-adapters`	-	Java, Playwright, and k6 publishing adapters

Runtime services

Service	Local URL	Notes
Platform Portal	http://localhost:8085	Main UI
Ingestion API	http://localhost:8081/swagger-ui.html	Result, project, requirement, integration, and TCM APIs
Analytics API	http://localhost:8082/swagger-ui.html	Flakiness, TIA, quality gates, alerts, logs
Integration API	http://localhost:8083/swagger-ui.html	Jira lifecycle service
AI API	http://localhost:8084/swagger-ui.html	Failure analysis and provider settings
Agent Hub	http://localhost:8087	Host port maps to container port `8086`
OpenSearch	http://localhost:9200	Log search backend
MinIO Console	http://localhost:9001	`platform` / `platform123`
PostgreSQL	localhost:5432	db `platform`, user `platform`, password `platform`
Kafka	localhost:9092	KRaft mode
Redis	localhost:6379	Agent checkpoints

The Grafana monitoring stack is not part of the default docker compose up. Start it with the override (see Full stack with monitoring) to also get Grafana (http://localhost:3000, admin/admin), Prometheus (http://localhost:9090, metrics + k6 remote-write), Loki (http://localhost:3100), and Promtail.

Quick start with Docker

No Java or Maven is required for the pre-built image path. The default docker-compose.yml is the full operational platform — a single docker compose up -d starts the data pipeline, storage, cache, log-search backend, and the LiteLLM gateway, runs Flyway through the db-migrate one-shot container, and then starts all backend services and the portal.

git clone https://github.com/ndviet/test-automation-platform.git
cd test-automation-platform

docker compose pull
docker compose up -d
docker compose ps

Stop the stack:

docker compose down

# Wipes database, Kafka, Redis, OpenSearch, and MinIO data.
docker compose down -v

Full stack with monitoring

docker-compose-full.yml is an additive override holding the self-contained extras: a bundled LiteLLM container (the default stack expects an external endpoint) and the Grafana observability stack (Grafana, Prometheus, Loki, Promtail). With the override present, platform-ai/agent default to the bundled litellm. Layer it on the default to run the full stack:

docker compose -f docker-compose.yml -f docker-compose-full.yml up -d
docker compose -f docker-compose.yml -f docker-compose-full.yml down   # stop it all

scripts/dev-up.sh --full (or COMPOSE_FULL=1 scripts/dev-up.sh) does the same with the bind-volume warmup.

Configuration

Create a .env file in the project root when you need non-default settings. Docker Compose reads it automatically.

# LiteLLM gateway — the single OpenAI-compatible endpoint the platform talks to
# (Claude/GPT/Gemini/etc. are routed by model id inside LiteLLM).
#
# The DEFAULT stack uses an EXTERNAL LiteLLM: platform-ai/agent start with these
# unset (empty), so the platform comes up from scratch without a gateway. Point them
# at your LiteLLM when ready — here, or per Org/Team/Project in the Portal (/settings/ai):
# LITELLM_BASE_URL=https://your-litellm/v1
# LITELLM_API_KEY=sk-...
#
# Prefer a bundled gateway? Run the full stack (-f docker-compose-full.yml) — it adds a
# `litellm` container that proxies to the Claude API, and platform-ai/agent then default to
# LITELLM_BASE_URL=http://litellm:4000/v1. Provide your Anthropic key (and optional proxy key):
ANTHROPIC_API_KEY=sk-ant-...
LITELLM_MASTER_KEY=sk-platform-local

# Ingestion auth, disabled by default for local development
API_KEY_AUTH_ENABLED=false

# Agent integrations
GITHUB_TOKEN=ghp_...
GITHUB_WEBHOOK_SECRET=...
JIRA_WEBHOOK_SECRET=...
LINEAR_WEBHOOK_SECRET=...
SLACK_BOT_TOKEN=xoxb-...
SLACK_SIGNING_SECRET=...
SLACK_APPROVAL_CHANNEL=#agent-approvals

Storage is selected with platform.storage.type:

Type	Use case
`filesystem`	Local service runs without object storage
`minio`	Docker Compose local stack
`s3`	AWS S3 or compatible production storage

Compose configures MinIO automatically and creates these buckets: platform-artifacts, platform-knowledge, platform-checkpoints, and platform-diffs.

Build from source

Prerequisites

Tool	Version
Java	21
Maven	3.9+
Docker + Docker Compose	27+
Node.js	20+ for portal and JS adapter development

Build all Java modules

mvn -B clean package -DskipTests

Build deployable services only

mvn -B clean package -DskipTests \
  -pl platform-ingestion,platform-analytics,platform-integration,platform-ai,platform-portal,platform-agent \
  -am

Run tests

mvn test

Build JS adapters

cd platform-adapters/js/playwright
npm ci
npm run build

cd ../k6
npm install --ignore-scripts
npm run bundle

Build and run local images

All service images build FROM a shared platform-base image (JRE + OS packages + non-root user + JVM entrypoint, defined in infrastructure/docker/base.Dockerfile). That common layer is built once instead of repeated in every service image.

Fastest path — build the base and all services in parallel with one command:

docker buildx bake          # builds platform-base:local + all six service images

Or with Compose (build the base once first, then the services):

docker buildx bake base                       # creates platform-base:local
docker compose up -d --build

CI builds the base once and publishes it to ghcr.io/<owner>/platform-base; the service images pull it via the BASE_IMAGE build-arg.

Run one service from source

Start infrastructure first:

docker compose up -d

Then run the service you are changing:

cd platform-ingestion
mvn spring-boot:run -Dspring-boot.run.profiles=local

cd platform-agent
mvn spring-boot:run -Dspring-boot.run.profiles=local

cd platform-portal/frontend
npm install
npm run dev

Submitting test results

HTTP upload

curl -X POST http://localhost:8081/api/v1/results/ingest \
  -F "file=@target/surefire-reports/TEST-example.xml" \
  -F "format=JUNIT_XML" \
  -F "teamId=my-team" \
  -F "projectId=my-project"

Supported formats:

JUNIT_XML, CUCUMBER_JSON, TESTNG, ALLURE, PLAYWRIGHT, NEWMAN, PLATFORM_NATIVE, K6, GATLING, JMETER.

Playwright adapter

# .npmrc
@ndviet:registry=https://npm.pkg.github.com
//npm.pkg.github.com/:_authToken=YOUR_GITHUB_TOKEN

npm install @ndviet/adapter-playwright

// playwright.config.ts
export default defineConfig({
  reporter: [
    ['list'],
    ['@ndviet/adapter-playwright', {
      endpoint: process.env.PLATFORM_URL,
      apiKey: process.env.PLATFORM_API_KEY,
      teamId: 'my-team',
      projectId: 'my-project',
    }],
  ],
})

k6 adapter

// load-test.js
import { platformHandleSummary } from './platform-k6-adapter.js'
export { platformHandleSummary as handleSummary }

k6 run \
  -e PLATFORM_URL=http://localhost:8081 \
  -e PLATFORM_API_KEY=plat_xxxx \
  -e PLATFORM_TEAM_ID=my-team \
  -e PLATFORM_PROJECT_ID=my-project \
  load-test.js

Java adapter and testkit

Use platform-adapter-java for existing suites that only need auto-publishing. Use platform-testkit-java for rich step tracking, tracing, retries, Playwright diagnostics, and native JSON publishing.

<dependency>
  <groupId>org.ndviet</groupId>
  <artifactId>platform-testkit-java</artifactId>
  <version>1.0.0-SNAPSHOT</version>
  <scope>test</scope>
</dependency>

See:

platform-adapters/java/USAGE.md
platform-testkit/java/USAGE.md
platform-adapters/js/playwright/README.md
platform-adapters/js/k6/README.md

Main APIs

Ingestion and platform management

Area	Endpoints
Results	`POST /api/v1/results/ingest`
Coverage	`POST /api/v1/coverage`
Teams	`GET/POST /api/v1/teams`, `PUT/DELETE /api/v1/teams/{id}`
Projects	`GET/POST /api/v1/projects`, `PUT/DELETE /api/v1/projects/{id}`
Executions	`GET /api/v1/projects/{projectId}/executions`, `GET /api/v1/executions/{runId}`
Requirements	`GET /api/v1/projects/{projectId}/requirements`, `GET /api/v1/projects/{projectId}/requirements/stats`
Integrations	`GET/POST /api/v1/projects/{projectId}/integrations`, `DELETE /api/v1/projects/{projectId}/integrations/{configId}`
API keys	`POST /api/v1/api-keys`, `GET /api/v1/api-keys`, `DELETE /api/v1/api-keys/{id}`
Test suites	`GET/POST /api/v1/projects/{projectId}/test-suites`
Test cases	`GET/POST /api/v1/projects/{projectId}/test-cases`, review and automation actions under `/test-cases/{tcId}`
Manual test runs	`GET/POST /api/v1/projects/{projectId}/test-runs`, execution updates under `/test-runs/{runId}`

Analytics

Area	Endpoints
Flakiness	`GET /api/v1/analytics/{projectId}/flakiness`, `POST /api/v1/analytics/{projectId}/flakiness/recompute`
Trends	`GET /api/v1/analytics/{projectId}/trends/pass-rate`, `/trends/mttr`, `/trends/duration`
Quality gates	`GET /api/v1/analytics/{projectId}/quality-gate`, `GET /api/v1/projects/{projectId}/quality-gate/ci`
Test Impact Analysis	`GET /api/v1/analytics/{projectId}/impact`, `GET /api/v1/analytics/{projectId}/impact/summary`
TIA trends	`GET /api/v1/analytics/tia/summary`, project TIA trend endpoints under `/api/v1/analytics/{projectId}/tia/*`
Alerts	`GET /api/v1/alerts/projects/{projectId}`, `GET /api/v1/alerts/org`
Release report	`GET /api/v1/projects/{projectId}/release-report`
Logs	`GET /api/v1/logs/runs/{runId}`, `GET /api/v1/logs/tests/{testId}`, `GET /api/v1/logs/runs`

AI and Agent Hub

Area	Endpoints
Failure analyses	`GET /api/v1/projects/{projectId}/analyses`, `POST /api/v1/projects/{projectId}/results/{resultId}/analyse`
AI settings	`GET/PUT /api/v1/ai/settings`, `POST /api/v1/ai/settings/test`
Agent workflows	`POST /hub/workflows`, `GET /hub/workflows`, `GET /hub/workflows/{workflowId}`
Node registry	`POST /hub/nodes/register`, `POST /hub/nodes/{nodeId}/heartbeat`, `GET /hub/nodes`
Test generation	`POST /hub/test-cases/{projectId}/generate`, `POST /hub/test-cases/{projectId}/{testCaseId}/generate-automation`
Impact analysis	`GET/POST /hub/impact/{projectId}`, `GET /hub/impact/{projectId}/prs`
Flaky healing	`POST /hub/healing/{projectId}/trigger`
Review decisions	`GET /hub/reviews`, `POST /hub/reviews/{reviewRequestId}/decide`
Webhooks	`/hub/webhooks/github`, `/hub/webhooks/jira`, `/hub/webhooks/linear`, `/hub/webhooks/slack/interactions`

Portal features

The React portal is served by platform-portal and backed by /api/portal/**.

Route	Purpose
`/`	Organization summary, teams, projects, recent quality signal
`/projects/:projectId`	Project quality detail
`/projects/:projectId/requirements`	Requirement inventory and stats
`/projects/:projectId/pr-analyses`	Agent PR analysis history
`/projects/:projectId/impact-analyses`	AI impact analysis workflow UI
`/projects/:projectId/settings`	Project, team, and integration configuration
`/projects/:projectId/test-cases`	Test suites, manual test cases, review, AI generation, automation trigger
`/projects/:projectId/test-runs`	Manual test runs and execution updates
`/projects/:projectId/flaky-tests`	Flakiness recompute and healing trigger
`/runs/:runId`	Ingested run detail
`/alerts`	Alert history
`/settings/api-keys`	Ingestion API key management
`/settings/ai`	LiteLLM gateway settings (base URL, key, models), test, and tool-config export

Observability

Signal	Where
Dashboards	Grafana at http://localhost:3000
Metrics	Prometheus at http://localhost:9090
Service logs	Loki through Grafana Explore
Test execution logs	OpenSearch indices `test_execution_logs-*`
k6 performance	Prometheus (remote write) plus Grafana dashboards
TIA trends	Grafana TIA dashboard and analytics APIs

Database migrations

Flyway migrations live in platform-core/src/main/resources/db/migration/ and currently run from V1 through V44.

They run automatically in Docker through the db-migrate service and are also available to Spring Boot services through the shared core module.

docker compose run --rm db-migrate

CI templates

The repository includes Test Impact Analysis helper templates under platform-ci-templates/:

GitHub Actions: platform-ci-templates/github-actions/test-impact-analysis.yml
Azure DevOps: platform-ci-templates/azure-devops/test-impact-analysis.yml
Jenkins: platform-ci-templates/jenkins/PlatformTIA.groovy

These templates calculate changed files, call the analytics TIA endpoints, emit Maven/Gradle test filters, and can fall back to full-suite execution based on risk.

Project structure

test-automation-platform/
|-- platform-common/              # Shared platform, agent, integration, and storage contracts
|-- platform-storage/             # Blob storage auto-config and implementations
|-- platform-core/                # JPA domain, repositories, Flyway migrations
|-- platform-ingestion/           # Result parsers, REST ingest, management, TCM APIs
|-- platform-analytics/           # Flakiness, trends, quality gates, TIA, alerts, log search
|-- platform-integration/         # Jira ticket lifecycle service
|-- platform-integration-config/  # Integration adapter registry and initial adapters
|-- platform-ai/                  # Failure classification and AI settings
|-- platform-agent/               # Agent Hub, workflows, nodes, webhooks, review gateway
|-- platform-portal/              # Spring Boot BFF and React SPA
|-- platform-testkit/             # Rich Java test instrumentation
|-- platform-adapters/            # Java, Playwright, and k6 adapters
|-- platform-ci-templates/        # TIA templates for CI systems
|-- infrastructure/               # Grafana, Prometheus, Loki, Promtail, Logstash, OpenSearch
|-- helm/                         # Kubernetes chart
`-- docker-compose.yml            # Full local stack

Name		Name	Last commit message	Last commit date
Latest commit History 61 Commits
.claude/skills		.claude/skills
.github		.github
docs		docs
helm/platform		helm/platform
infrastructure		infrastructure
platform-adapters		platform-adapters
platform-agent		platform-agent
platform-ai		platform-ai
platform-analytics		platform-analytics
platform-ci-templates		platform-ci-templates
platform-common		platform-common
platform-core		platform-core
platform-ingestion		platform-ingestion
platform-integration-config		platform-integration-config
platform-integration		platform-integration
platform-llm		platform-llm
platform-portal		platform-portal
platform-security		platform-security
platform-storage		platform-storage
platform-testkit		platform-testkit
scripts		scripts
spec		spec
tasks		tasks
.dockerignore		.dockerignore
.gitignore		.gitignore
COST.md		COST.md
PLATFORM_PLAN.md		PLATFORM_PLAN.md
PLATFORM_PLAN_PHASE2.md		PLATFORM_PLAN_PHASE2.md
README.md		README.md
SPEC.md		SPEC.md
docker-bake.hcl		docker-bake.hcl
docker-compose-full.yml		docker-compose-full.yml
docker-compose.override.yml		docker-compose.override.yml
docker-compose.yml		docker-compose.yml
pom.xml		pom.xml

Uh oh!

Folders and files

Latest commit

History

Repository files navigation

Test Automation Platform

Demo

What's in v2.0

Architecture

Maven modules

Runtime services

Quick start with Docker

Full stack with monitoring

Configuration

Build from source

Prerequisites

Build all Java modules

Build deployable services only

Run tests

Build JS adapters

Build and run local images

Run one service from source

Submitting test results

HTTP upload

Playwright adapter

k6 adapter

Java adapter and testkit

Main APIs

Ingestion and platform management

Analytics

AI and Agent Hub

Portal features

Observability

Database migrations

CI templates

Project structure

About

Resources

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Uh oh!

Contributors

Uh oh!

Languages

Packages