Skip to content

Releases: jasonzhang06-source/sharelint

ShareLint 1.0.0

Choose a tag to compare

@jasonzhang06-source jasonzhang06-source released this 07 Sep 01:56

ShareLint 1.0.0

The first stable release of ShareLint's local privacy preflight. This release
focuses on a smoother first run, cross-platform delivery, and maintenance of the
existing scanner. Detection scope and machine-readable contracts remain unchanged.

Improvements

  • Clearer command help and an explicit synthetic-data banner for sharelint demo.
  • Console and HTML reports show REVIEW for findings below the blocking threshold,
    while preserving JSON/SARIF verdicts and exit codes.
  • Clear separation between the shareable archive and private reports/receipts.
  • Corrected the pack help example and clarified that incomplete PDF/image coverage
    blocks packing; unknown rule errors do not echo terminal control sequences.
  • A documented 1.x compatibility policy and updated English/Chinese guidance.

Downloads

The release includes standalone archives for Linux glibc x86-64, Windows x86-64,
Intel macOS, and Apple silicon macOS. They include Python and need no Python
installation on the destination computer. Each archive has a .sha256 file and
GitHub build provenance. Native builds and their final archives are smoke-tested
before publication.

Windows binaries are not Authenticode-signed. macOS binaries are ad-hoc signed,
without Developer ID signing or notarization. See the
platform guide.

Install or upgrade

pipx install sharelint==1.0.0
sharelint demo

To select this exact release in an existing pipx installation: pipx install --force sharelint==1.0.0.

With uv: uvx --from sharelint==1.0.0 sharelint demo, or
uv tool install sharelint==1.0.0 for a persistent installation.

PyPI ·
Changelog

Compatibility and scope

Python 3.11+; existing commands, exit codes, report schema 1.0.0, receipt schema
1.0.0, and SARIF 2.1.0 remain compatible. There are no third-party runtime Python
package dependencies. PDF pages and image pixels are not OCR-scanned, originals
are not cleaned or modified, and incomplete coverage still blocks pack.
A pass describes the configured checks, not a guarantee that a file is safe.

ShareLint 0.1.1 — smoother first run

Choose a tag to compare

@jasonzhang06-source jasonzhang06-source released this 24 Aug 04:25

Highlights

  • Added sharelint demo --report for safely writing local HTML, JSON, SARIF,
    or console demo reports.
  • Made isolated installation with pipx or uv the primary path.
  • Added installation troubleshooting, architecture documentation, and a
    unified documentation index.
  • Improved English and Chinese first-run guidance and contribution paths.

Install

  • Install: pipx install sharelint==0.1.1
  • Try the demo: sharelint demo
  • Select this version: pipx install --force sharelint==0.1.1
  • Install with uv: uv tool install sharelint==0.1.1
  • Generate local HTML: sharelint demo --format html --report sharelint-demo.html

PyPI: pypi.org/project/sharelint

Alpha limitations

ShareLint does not currently OCR PDF pages or image pixels, sanitize original
files, or guarantee that a passing artifact is safe or compliant.

ShareLint 0.1.0 — Scan before you share

Choose a tag to compare

@jasonzhang06-source jasonzhang06-source released this 24 Aug 03:13

ShareLint 0.1.0 (Alpha)

Scan before you share.

ShareLint is a local, fail-closed privacy preflight for files, folders, and
nested archives.

Highlights

  • Scan directories, ZIP archives, Office documents, PDFs, images, and text
    locally.
  • Detect secrets, PII, metadata, speaker notes, hidden sheets, and embedded
    objects.
  • Preserve complete provenance through nested containers.
  • Hide matched values in console, JSON, SARIF, and HTML reports.
  • Expose partial and unsupported coverage instead of reporting a false clean
    result.
  • Create deterministic share bundles with output rescanning and hash-bound
    receipts.
  • Zero third-party runtime dependencies.

Install

  • python -m pip install sharelint==0.1.0
  • sharelint demo

Alpha limitations

PDF pages and image pixels are not currently OCR-scanned. A passing result is
a policy and coverage result, not a guarantee that an artifact is safe or
anonymous.