AGENTS.md

Benchmark Scripts

Overview

This folder is the canonical home for benchmark, baseline, and benchmark-data curation entrypoints. If a benchmark path in docs or CI disagrees with this folder, this folder wins and compatibility wrappers should be updated.

Key Components

• benchmark_*.py: benchmark runners and regression gates.
• baseline_trivial.py: cheapest-enemy benchmark.
• curate_500_amp_benchmark.py: benchmark dataset curation.
• expand_benchmark.py: benchmark expansion utility.

Diagrams (Mermaid)

• Flowchart

flowchart TD
  Inputs["Toy benchmark inputs"] --> Runner["Benchmark script"]
  Runner --> Metrics["JSON or console metrics"]
  Metrics --> Docs["docs/METRICS_CURRENT.md"]
  Metrics --> Gates["CI / reviewer gate"]

Loading

• Component Diagram

flowchart LR
  BenchScripts["scripts/benchmarks"] --> BenchPkg["src/openamp_foundry/benchmark"]
  BenchScripts --> Features["src/openamp_foundry/features"]
  BenchScripts --> Outputs["outputs/*.json"]
  Tests["tests/benchmarks"] --> BenchScripts

Loading

• Sequence Diagram

sequenceDiagram
  participant Dev as Contributor
  participant Script as Benchmark script
  participant Code as Benchmark code
  participant Output as Metrics artifact
  Dev->>Script: run benchmark
  Script->>Code: compute metrics
  Code-->>Script: results
  Script->>Output: write JSON/report
  Output-->>Dev: reviewable evidence

Loading