CI Evidence Pack
Version updated for https://github.com/OrygnsCode/ci-evidence-pack to version v1.0.0.
- This action is used across all versions by ? repositories.
Go to the GitHub Marketplace to find the latest changes.
Action Summary
The CI Evidence Pack Generator is a GitHub Action and CLI tool designed to create secure, deterministic evidence bundles for compliance and audit purposes. It automates the collection of source code context, build artifacts, dependencies, and SBOMs into a signed, verifiable tarball, providing auditable proof of software builds for regulated environments (e.g., SOC 2, ISO). The tool ensures security through keyless signing, strict verification of manifests and signatures, and comprehensive metadata collection.
Release notes
CI Evidence Pack Generator v1.0.0
First stable release of ci-evidence-pack: a CLI + GitHub Action that generates a deterministic, read-only CI evidence bundle for audits/compliance.
Highlights
- Generates a single
.tar.gzevidence bundle containing:- Git and CI run metadata
- Dependency snapshot (
pip_freeze.txtwhen Python project detected) - SBOM output (CycloneDX via
cyclonedx-pywhen available; Syft optional) - User-included build artifacts / reports
- A strict SHA256 manifest for integrity verification
- Verification built-in:
ci-evidence-pack verifyenforces strict structure, blocks unsafe paths, and checks hashes. - Optional signing: supports keyless signing via Sigstore Cosign in GitHub Actions (OIDC).
GitHub Action
- The repository is publishable as a GitHub Action (root
action.yml). - Recommended workflow permissions when signing:
contents: readid-token: write
Notes on determinism
Output is bit-for-bit reproducible when inputs are held constant (included files, SOURCE_DATE_EPOCH, and captured CI env vars such as GITHUB_RUN_ID).
How to use
- CLI:
ci-evidence-pack packci-evidence-pack verify dist/ci-evidence-pack_*.tar.gz
- GitHub Actions:
uses: OrygnsCode/ci-evidence-pack@v1.0.0
Compatibility
- Python package published as
ci-evidence-pack - Designed for CI environments and local runs
If you hit any issues, open an issue with your workflow YAML + the --json output from pack or verify.