Performance Logic

February 7, 2026 ยท View on GitHub

Performance Logic

Establishes performance goals and benchmarking strategies for the Docgraph workspace validation.

Decision

We use criterion for micro-benchmarks and set performance targets:

  • Linting: < 50ms for typical workspaces
  • Graph Generation: Scalable to 1000+ nodes

Rationale

  • Benchmarks located in benches/ directory
  • Performance tests run in CI/CD pipeline
  • Critical paths in core logic are optimized

Context

Performance is critical for recursive graph validation. Users expect fast feedback during development.