BC Observatoryunofficial, made by waldo
community postAL Developmentcommunity - not Microsoftreviewed - checked by Opus

AL Runner: Run AL Unit Tests Without a BC Service Tier

AL Runner is a CLI tool that transpiles AL code to C# and executes unit tests in-memory without a BC service tier, Docker, or SQL Server, completing test runs in seconds instead of 20-30 minutes. It uses the BC compiler's public API to convert AL objects to C#, rewrites runtime types with mocks, and compiles everything with Roslyn for fast local testing.

Open in Claudehttps-stefanmaron-com-posts-al-runner-run-al-tests-without-bc--693c9fd16b.md

post Stefan Maron

Stefan Maron · 2026-04-24 · 2,128 words · about 10 min

"AL Runner drops that overhead entirely. Transpilation, in-memory compilation, and test execution all happen in seconds."
Show the post hereRead on Stefan Maron

Key points

  • AL Runner transpiles AL to C# using Microsoft.Dynamics.Nav.CodeAnalysis.Compilation.Emit(), rewrites BC runtime types with in-memory mocks, and executes tests via Roslyn compilation without a service tier.
  • Test execution runs in seconds compared to 15-30 minute container-based pipelines, making it ideal for CI pre-checks and AI coding agents that need fast feedback loops.
  • Dependencies are handled via auto-stubs, hand-written stubs, or compiled DLLs; code inside .app packages requires stubs or –compile-dep workaround.
  • Supports codeunits, tables, enums, queries, record operations, events, TestPage, JSON types, and many AL features; does not support transaction semantics, parallel sessions, UI rendering, or real HTTP I/O.
  • Available as a global dotnet tool with cross-platform support (Windows, Linux, macOS), automated BC Service Tier DLL downloads, and machine-readable output for agent integration.

Quotes

  • “AL Runner drops that overhead entirely. Transpilation, in-memory compilation, and test execution all happen in seconds.” (Demonstrates the core value: eliminating the 15-30 minute container startup time that slows down development cycles.)
  • “AL Runner is a pure CLI tool - cross-platform, no GUI, machine-readable output. That makes it a natural fit for coding agents.” (Shows how the design intentionally supports AI agent integration for automated test-driven development workflows.)

Context

  • Features: AL unit test execution, In-memory test compilation, Dependency stubbing, TestPage support, Event subscriber testing, Debug Adapter Protocol, CI/CD integration, Machine-readable output
  • Versions: BC 26, BC 27, BC 28

Source: Stefan Maron, community blog. Summary, key points and quotes are derived (CONTENT-NOTICE.md); read the original for the full text.