← ClaudeAtlas

lidguard-release-validationlisted

LidGuard validation and release workflow reference. Use when working on build validation guidance, current validation scope, missing work, automated verification, .NET tool packaging, NuGet publishing, package commands, or local install smoke tests.
airtaxi/LidGuard · ★ 7 · AI & Automation · score 76
Install: claude install-skill airtaxi/LidGuard
# LidGuard Release Validation ## Build Validation Note - Local build, test, publish, pack, and reinstall validation commands can fail once because of transient Windows Defender file-lock interference. - Retry the same validation command before taking any broader recovery action; when this specific issue is the cause, a retry is typically enough. - Do not bring down any build server just because the first validation attempt failed with this known Defender issue. - When validating multiple Runtime Identifiers locally from the same worktree, run the commands sequentially rather than in parallel. Concurrent `dotnet build` or restore activity against the same project can race on shared `obj` artifacts and produce false-negative reference or restore failures that disappear on a clean sequential rerun. ## Current Validation Scope - Current manual runtime validation has only covered Windows with Codex. - Windows, Linux, and macOS RID compile validation can catch platform compilation regressions, but it is not a substitute for runtime behavior validation on each target OS. - Linux systemd/logind runtime behavior, macOS runtime behavior, Claude Code hooks, GitHub Copilot CLI hooks, OpenCode hooks, Provider MCP flows, and cross-provider concurrent-session behavior still need real environment validation before being treated as verified. ## Missing Work The Windows, Linux, and macOS CLI hook receiving path is implemented for Codex, Claude Code, GitHub Copilot CLI, and OpenCode. Rema