Skip to content

Security: arcusis/shaar

SECURITY.md

Security policy

Shaar is under active development. No fixed support window or security update service is currently promised. The bundled Chromium engine is pinned; a pin does not provide ongoing security updates.

Report a vulnerability

Report suspected vulnerabilities privately through GitHub Security Advisories. Include affected version or commit, impact, and reproducible steps when safe. Do not open a public issue for an unpatched vulnerability.

The maintainers will review reports and coordinate a response through the advisory. Do not include secrets or unrelated personal information in reports.

Accepted Scorecard findings

These OpenSSF Scorecard results stay short of a perfect score on purpose. Each one is a posture signal about project process. Closing one by adding a tool whose only job is to move the number is out of scope for this release. GitHub secret scanning and push protection stay off.

Fuzzing. Shaar does not yet have a parser of its own that a fuzzer should target. The pinned Chromium engine is already fuzzed by the Chromium project. A placeholder fuzz target in this repository would not exercise Shaar.

CII Best Practices. The project has no OpenSSF Best Practices badge. Publishing a badge before the criteria are met, including required review and project age, would not make the release safer.

Maintained. Scorecard scores this at 0 until a repository is 90 days old. This repository was created on 2026-10-24, so the score cannot rise before 2027-01-22 no matter how often the code changes. Development is active.

Code review. Recent production commits do not have a second person's approval. During this release sprint, ruleset 26350516 stays disabled. The checked-in ruleset requires the CI status check and the merge queue. Pull-request review is a separate setting, and this ruleset leaves it unset. Turning the ruleset back on remains a release action for the repository owner.

SAST and CI tests. CodeQL for Swift, Python, and Actions, and the pull-request test workflows, are already running. Scorecard gives both checks 9 because 28 of the last 30 commits were covered. The missing two are historical. A second scanner would not close that gap.

There aren't any published security advisories