Skip to content

v0.6.1

Latest
Latest

Choose a tag to compare

View all tags
@github-actions github-actions released this 25 May 14:51
v0.6.1
58ebc4e

Release Notes

v0.6.1

UI/UX redesign release: refreshed the Wails desktop, local web controller, and standalone HTML report surfaces using the new Quietscope security-dashboard system from the redesign references.

Changed

  • Refreshed the Wails desktop start, running, completed, findings, AI Control Center, and backup-first preview flows.
  • Refreshed the local web controller with a darker loopback dashboard, clearer audit phases, locked running controls, and safer preview/edit modal language.
  • Refreshed the standalone HTML report dashboard, findings, AI hardening, AI inventory/catalog, control center, cleanup, and privacy masking surfaces.
  • Version bumped to v0.6.1.

Security

  • Local-server report actions now require loopback hosting plus a session token.
  • Action API requests are validated against the selected report artifact and action availability metadata, so disabled UI actions are also blocked server-side.
  • Report and desktop log rendering received additional escaping around user-controlled audit values.

v0.6.0

AI Control Center release: manageable local AI artifacts are now visible across JSON, HTML, local web, and desktop surfaces with preview/diff/backup-first actions.

Added

  • manageable_artifacts report contract covering skills/guides/rules/prompts, MCP server entries, caches/logs, local models, and action availability metadata.
  • Safe remediation core v2 with PreviewAction, ExecuteAction, CreateBackup, RestoreBackup, ReadArtifact, and SaveArtifact.
  • Structural MCP manager support for JSON, TOML, and YAML configs, including mcpServers, servers, command, args, env key metadata, URL fields, serverUrl, and transport/type fields.
  • Local web API endpoints: GET /api/artifacts, POST /api/actions/preview, POST /api/actions/execute, and POST /api/actions/restore.
  • Wails bridge parity methods: ListManageableArtifacts, PreviewAction, ExecuteAction, RestoreBackup, ReadArtifact, and SaveArtifact.
  • AI Control Center sections in the static HTML report, local browser UI, and Wails desktop preview.

Changed

  • Static HTML reports now show action buttons disabled with honest reasons instead of implying offline mutation support.
  • Cache/log cleanup and model storage are split: allowlisted caches can be cleaned with backup; model directories are manual-only by default.
  • Gemini, Google Antigravity, .agents, .gemini, Hermes, OpenCode, and additional MCP path coverage are included in management metadata.
  • Version bumped to v0.6.0.

Security

  • Write/delete/disable/fix/restore actions require allowlist checks, symlink blocking, and a backup before mutation.
  • Secret paths, Keychain data, .env files, SSH/private keys, and cloud credential roots remain blocked from read/edit workflows.
  • MCP commands/scripts/hooks are never executed; config mutation is parser-based rather than string replacement.

v0.5.1

UI layout polish and release asset cleanup for the desktop/report refresh.

Fixed

  • Standalone HTML reports no longer overflow horizontally at desktop widths when the sidebar is fixed.
  • Findings tables now stay constrained to the report content area, with long content wrapping inside table cells.
  • The report risk gauge initializes on first render instead of showing an empty placeholder state.
  • The README logo asset is now a valid PNG file matching its .png extension.

Changed

  • Version bumped to v0.5.1.

v0.5.0

Desktop UI & Report overhaul with full-screen Audit Findings Explorer.

Added

  • Full-screen Audit Findings Explorer: dual-pane interactive workstation with scrollable finding cards (left) and full remediation details panel (right) including evidence, detection commands, auto-fix status, reclaimable size, and copy-to-clipboard actions.
  • Horizontal Metrics Summary Bar: compact inline bar displaying Risk Score, Findings count, AI Risks, Secrets exposure, Duration, and "Open In Browser" action — visible at a glance in the findings view.
  • Active Scan Pulse Loaders: animated pulsing skeleton indicators for Findings, AI Risks, and Secrets metrics while audits are running or queued, preventing misleading zero values.
  • Analyzing Gauge State: circular risk gauge enters a pulsing "Analyzing..." mode with a blue ring during active scans instead of prematurely showing "Secure" with score 0.
  • XSS Protection: all finding data is escaped via escapeHTML() before rendering in innerHTML templates.
  • Standalone HTML Report redesign: interactive glassmorphic safety dashboard with animated SVG risk gauge, privacy masking toggle, remediation simulator, coordinated multi-faceted filters, and print-optimized PDF export styling.

Fixed

  • Duration calculation bug: Go's zero-value timestamps (0001-01-01T00:00:00Z) no longer produce -63915244578.8s — safely detected and rendered as "-" for queued jobs or live count-up for running jobs.
  • Metrics display during active scans: Findings, AI Risks, and Secrets no longer show misleading 0 while scan is in progress.

Changed

  • Version bumped to v0.5.0.
  • Desktop UI upgraded from basic table-based findings viewer to professional dual-pane security analysis workstation.
  • Wails desktop header badge updated to v0.5.0.
  • SECURITY.md supported versions updated (v0.5.x active, v0.4.x maintenance).

Known limitations

  • Wails UI remains in developer preview and requires Wails CLI framework installation to compile.
  • Cleanup operations remain CLI-only for safety.

v0.4.0

Wails Desktop Preview and Basic Windows Audit release.

Added

  • Wails desktop frontend scaffolding under desktop/ providing a native window wrapper around Quietscope audit engine.
  • Real-time progress and logs streaming from Go core to Wails UI via Wails Events bus.
  • Basic Windows audit module with persistence checks for User/System Startup folders, registry Run keys, and User PowerShell profiles.
  • Basic Windows security checks including SSH file permissions, Windows Defender status, Firewall profiles status, and User Account Control (UAC) EnableLUA status.
  • Integration tests and mock fixtures for Windows checks validation.

Changed

  • Version bumped to v0.4.0.
  • README updated to reflect Wails developer preview availability and new basic Windows support levels.
  • CLI main registry gates now load Windows audit modules on Windows hosts.

Known limitations

  • Wails UI is in active developer preview and requires Wails CLI framework installation to compile.
  • UI does not support destructive cache cleanup operations (kept CLI-only for safety).
  • Windows checks are read-only and metadata-first; they do not cover active malware removal or deep registry hardening.

v0.3.0

Initial Linux audit module release.

Added

  • Linux check registry that enables linux_persistence and linux_security groups on Linux only.
  • Linux persistence metadata checks for systemd system/user units, cron paths, desktop autostart entries, and Linux shell startup files.
  • Safe sampled parsing of systemd, cron, and desktop autostart startup lines with redaction and command/network risk flags.
  • Linux security basics for user/system SSH metadata, host key permissions, firewall status, update metadata timestamps, and sudoers metadata.
  • Linux smoke test that builds the CLI and generates TXT/JSON/HTML reports on Ubuntu.
  • Fixture-driven unit tests for Linux systemd/cron/autostart metadata, firewall parsing, unavailable firewall status, SSH/update/sudoers metadata, and path handling.

Changed

  • Version bumped to v0.3.0.
  • README platform support now marks Linux as an initial audit platform instead of soft-supported only.
  • CI now includes a Linux smoke job alongside macOS smoke and cross-compile validation.

Known limitations

  • Linux checks are conservative and metadata-first. They do not claim full distro hardening coverage.
  • Firewall ruleset interpretation is best-effort and should be manually verified.
  • Package update checks do not run network package-manager commands.
  • Windows remains soft-supported with full Windows audit work planned later.

v0.2.0

Soft cross-platform CLI release.

Added

  • Platform capability helpers for Darwin, Linux, Windows, macOS security settings, launchd persistence, Linux systemd, Windows registry, and common filesystem checks.
  • SKIPPED finding status so unsupported OS-specific checks are visible without being treated as failures.
  • Linux and Windows safe metadata roots for common AI/MCP config discovery.
  • Platform-specific file ownership handling so Windows builds can represent unavailable UID/GID values as 0.
  • CI matrix for Ubuntu, macOS, and Windows, plus cross-compile validation for release targets.
  • Tag-triggered release workflow that builds Darwin arm64/amd64, Linux amd64, Windows amd64, and SHA256SUMS.
  • Platform support table and build examples for macOS, Linux, and Windows.

Changed

  • Version bumped to v0.2.0.
  • macOS-only security, update, launchd persistence, launchd permission, Time Machine, and macOS system metadata checks are gated by platform support.
  • Linux and Windows runs keep common checks and reports available while explaining unsupported macOS checks in plain language.
  • Report summaries include a skipped count.

Known limitations

  • macOS remains the only fully supported audit platform in v0.2.
  • Linux and Windows are soft-supported: common checks and report generation are best-effort, not full platform audits.
  • Linux systemd/firewall/package-manager audit checks are planned for v0.3.
  • Windows Defender/Firewall/startup/registry audit checks are planned after the Linux module work begins.
  • Cross-compilation may require a Go toolchain installation that includes the target standard libraries.

v0.1.1

Terminal observability and local UI control release.

Added

  • Short terminal process log with a compact progress bar for audit checks, report writing, cleanup dry-runs, and local server startup.
  • --ui mode for a local 127.0.0.1 audit control UI.
  • UI audit management: configure report outputs, deep checks, AI audit checks, no-sudo mode, cleanup dry-run, output directory, project root, and max scan size.
  • UI run history with live progress events, summary metrics, cancel action, and generated report links.
  • Same-origin UI token for local mutating API requests.
  • README documentation for CLI, self-contained report UI, and audit control UI modes.

Changed

  • Version bumped to v0.1.1.
  • The local UI notice now reflects that the controller can launch local audits but still never uploads report or file contents.

v0.1.0

Initial open-source-ready release.

Added

  • Go-first single-binary CLI.
  • Defensive local macOS audit checks for system info, security settings, persistence, permissions, storage, cleanup candidates, AI tooling, MCP configs, local LLM listeners, prompt-injection artifacts, and secret metadata.
  • AI Skills & Context Inventory for Claude, Codex, Cursor, Copilot, VS Code, Continue, Cline, Roo, Windsurf, Aider, Open Interpreter, and generic AI-agent artifacts.
  • AI-related directory disk usage inventory with context impact scoring and manual-review recommendations for model directories.
  • Extended AI Tool Catalog covering AI IDEs, coding agents, CLI agents, MCP clients/servers, local LLM runtimes, desktop wrappers, AI security scanners, Hermes Agent, OpenCode/opencode, and hosted/cloud agent local artifacts.
  • Additional AI Models & Providers inventory for Qwen, DeepSeek, Kimi/Moonshot, GLM/Z.ai/Zhipu, MiniMax, Doubao/ByteDance, ERNIE/Baidu, Baichuan, Yi/01.AI, InternLM, Hunyuan/Tencent, StepFun, SenseNova, and related local caches/provider configs.
  • TXT and JSON reports by default.
  • Optional self-contained HTML UI with dashboard, filters, privacy mode, AI section, cleanup section, print, and copy-summary.
  • Safe cleanup dry-run and exact confirmation phrase enforcement.
  • No external dependencies, telemetry, analytics, auto-update, or remote report upload.
  • GitHub Actions CI, smoke test, issue templates, PR template, SECURITY, CONTRIBUTING, CODE_OF_CONDUCT, and MIT license.

Known limitations

  • TCC permissions are reported as manual review metadata in v0.1.0.
  • Software update availability is not automatically queried to preserve local-only defaults.
  • Binary plist parsing is best-effort without external dependencies.
  • MCP schema coverage is best-effort and will expand over time.
Assets 7
Loading