Incremental scan cache

vulkro-sf antipatterns caches each file's findings, so the next scan only re-runs detectors on files whose source changed.

vulkro-sf antipatterns .                                # uses ~/.vulkro/sf-antipattern-cache.json
vulkro-sf antipatterns . --cache-dir .vulkro-cache      # cache in the workspace (CI)
vulkro-sf antipatterns . --no-cache                     # full re-scan

Setting	Flag	Variable
Cache folder (default ~/.vulkro/)	--cache-dir <path>	VULKRO_SF_CACHE_DIR
Turn the cache off	--no-cache	VULKRO_SF_NO_CACHE=1

What invalidates an entry

Each entry is keyed by the file's SHA-256, the detector version, the --compliance profiles and the project's own namespaces (from sfdx-project.json, --exclude-namespace and .vulkro-sf.yml). Changing any of these re-runs the affected files. A release that changes a detector re-runs every file.

What is cached

The findings before filtering. So disable in .vulkro-sf.yml and // vulkro: ignore AP-XXX comments are applied on every read: removing either brings the findings back on the next scan without a re-scan.

The cache file is plain JSON. Delete it to force a full re-scan.

In CI

Keep the cache folder between runs:

- uses: actions/cache@v4
  with:
    path: .vulkro-cache
    key: vulkro-sf-${{ github.run_id }}
    restore-keys: vulkro-sf-
- run: vulkro-sf antipatterns . --cache-dir .vulkro-cache --format junit > junit.xml

See also: CI/CD