← All benchmarks
May 2026
Angular + NgRx · ngrx-platform
Official NgRx monorepo — Angular 18, Nx workspace, ~1,379 TypeScript files, 248 spec files. Includes @ngrx/store, @ngrx/effects, @ngrx/entity, @ngrx/router-store, @ngrx/component-store.
View on GitHub ↗Benchmark setup
Results summary
Conservative local comparison
8 / 10 → 9 / 10 +1 pt
The raw no-Tessra baseline reached 9/10, but we apply a conservative -1 adjustment because the editor/chat may have retained benchmark context. Tessra's verified run reached 9/10 and reduced manual exploration through symbol-level navigation, caller/callee context, and grouped architecture context. ~27 files avoided · 9 symbols with context
Results per case
| Case | Question | Conservative baseline | With Tessra | What Tessra surfaced |
|---|---|---|---|---|
| 01 | When provideEffects() is called, which internal NgRx classes are instantiated and in what order? | 2 / 2 | 2 / 2 | Effects DI flow, caller/callee context, instantiation order |
| 02 | List every spec file that directly tests symbols from effect_creator.ts, and list the test names. | 2 / 2 | 2 / 2 | Direct spec file, test names, related test coverage |
| 03 | List all Nx library projects in this workspace, their Nx tags, and direct dependencies. | 1 / 2 | 1 / 2 | Limited: Nx metadata lives in config files, not the symbol graph |
| 04 | What methods does EntityStateAdapter define internally, not the public EntityAdapter interface? | 1 / 2 | 2 / 2 | Interface vs implementation boundaries, adapter internals, state operator helpers |
| 05 | For router-store: list all NgRx actions defined, their reducer, and the exact state type produced. | 2 / 2 | 2 / 2 | Actions, reducer behavior, pass-through actions, state type |
Key findings
Conservative lift, stronger navigation
The clean Tessra run reached 9/10 on local NgRx code. The raw no-Tessra baseline also reached 9/10, but we report a conservative 8/10 baseline because the editor/chat may have retained benchmark context. With that adjustment, Tessra shows a +1 point lift and clear workflow value: faster symbol navigation, caller/callee context, related specs, interface boundaries, reducer behavior, and state types with less manual exploration.
The win is not only the extra point. The bigger win is a cleaner path to the engineering answer.
Known limitations
This is a conservative comparison. The raw no-Tessra baseline reached 9/10, but we apply a -1 adjustment because prior editor/chat context may have influenced the result. This benchmark should not be read as a 0% → 100% claim. It measures implementation tracing and workflow leverage on a public, well-structured NgRx monorepo.
Try it yourself
See what Tessra surfaces in your repo.
Index an Angular, Django, or Flutter repo and try local context for 7 days.