Files
git.stella-ops.org/docs/features/dropped/multi-scanner-comparative-benchmarking.md

24 lines
1.3 KiB
Markdown

# Multi-scanner Comparative Benchmarking
## Status
NOT_FOUND
## Description
Advisory describes a benchmarking protocol comparing StellaOps scan results against Trivy/Grype/Snyk with precision/recall metrics. No CLI comparison tool or benchmark harness found.
## Why Not Implemented
- No dedicated CLI comparison tool or multi-scanner benchmark harness found
- A `compare.py` script exists at `src/__Tests/__Benchmarks/tools/compare.py` but it appears to be a general comparison utility, not a full multi-scanner benchmarking protocol
- The Bench module (`src/Bench/`) has benchmarking infrastructure (LinkNotMerge scenarios, Prometheus reporting) but not scanner comparison harnesses
- Golden corpus exists at `src/__Tests/__Benchmarks/golden-corpus/` with VEX scenarios and severity levels, which could serve as ground truth for scanner comparison
- The Scanner module has its own benchmark and test infrastructure but does not compare against external scanners (Trivy/Grype/Snyk)
- This would require external scanner integration which conflicts with the offline-first posture
## Source
- Feature matrix scan
## Notes
- Module: Bench
- Modules referenced: N/A
- Related: `src/__Tests/__Benchmarks/tools/compare.py` (comparison utility), `src/__Tests/__Benchmarks/golden-corpus/` (ground truth data)