Agent Fleet
Every agent runs on LangGraph, served through Core42 Compass. Each card shows current state, autonomy ceiling, success rate and the typed tools it can call.
Scope A — Agentic QA Suite
8 agentsReads JIRA stories and Confluence design to generate risk-based test cases (functional, negative, boundary, regulatory).
Generates compliant synthetic test data (Emirates ID, IBAN check digits, account profiles) and provisions datasets. Synthetic only.
Produces pytest, Playwright, Appium and WinAppDriver scripts; extends the existing ~150-test suite.
Runs suites via Azure Pipelines; clusters failures; triages flaky vs real regressions; opens JIRA defects.
Generates and runs JMeter/K6 scripts from NFRs; baseline and trend analysis.
Integrates SonarQube (SAST) and Qualys (DAST) into pipelines; correlates and deduplicates findings.
Produces Allure-style and executive reports; updates JIRA dashboards.
Runs minimal customer-mimicking checks in PROD after deployment; produces go/no-go evidence.
Scope B — Requirement Analysis
4 agentsIngests a product concept brief from Confluence/JIRA into a structured concept summary.
Concept + existing Epics/Stories (RAG) → draft Epics, User Stories, Acceptance Criteria in JIRA.
Solution Design doc in Confluence: traceability matrix, NFRs, integration agreements, architecture & sequence diagrams.
Reviews all artifacts → gaps, ambiguities, inconsistencies, and open questions for humans.
Scope B — Developer Agents
9 agentsiOS + Android builds, Jest unit tests. Delivers as PR to Azure Repos, gated by SonarQube + human reviewer.
Node build, Jest/Vitest unit tests. PR to Azure Repos.
dotnet build & test. PR to Azure Repos.
poetry/pip, pytest. PR to Azure Repos.
Headless CMS components. PR to Azure Repos.
Maven build of MuleSoft apps. PR to Azure Repos.
C# plugin, Power Platform CLI. PR to Azure Repos.
X++/.NET extensions. PR to Azure Repos.
Flow definitions as JSON / managed solutions. PR to Azure Repos.