Laravel Toolkit Laravel Package

Product Decisions This Supports

• Developer velocity vs. CI cost trade-off. The headline capability — inheriting a TIA (Test Impact Analysis) baseline across git worktrees and branches — directly attacks the "first test run and final preflight re-run everything" problem. That's a measurable decision input: faster inner loop, lower CI minutes, quicker feedback on every ticket branch. Use it to justify or retire parallel tooling, custom scripts, and bespoke cache-copying hacks in consuming apps.
• Build vs. buy on internal tooling. This release explicitly absorbs logic that consuming apps were carrying themselves ("consuming apps stop carrying their own copy of it in scripts"). If a team maintains homegrown shell scripts around Pest caching, preflight, retries, or static analysis, this package is a credible replacement — and the maintenance burden shifts off your team.
• Test-gate policy decisions. toolkit:preflight now refuses a coverage gate that silently disables TIA (e.g. --exclude-testsuite=Browser). That converts a quiet, expensive footgun into a hard failure. Good input for a policy decision on what "green" means in CI, and for standardizing preflight across repos.
• Scope of "any base branch," not just main. Support for arbitrary base branches means this scales to release branches, hotfix branches, and long-lived feature lines. Relevant if your roadmap includes trunk-based variants, release trains, or multi-version support.
• Platform/standardization roadmap. A named toolkit:* command family (tia, preflight, retry) plus a guard hook suggests an emerging internal standard for Laravel project hygiene. Adopting it is a step toward consistency across your portfolio of Laravel services.
• Concrete use cases to pilot: (1) a single repo with worktree-based ticket workflow, measuring wall-clock and CI-minute deltas; (2) a monorepo-adjacent set of Laravel apps sharing a preflight contract; (3) any team where pest --baseline was being blocked by a hook and silently degrading.

When to Consider This Package

Consider adoption when:

• Your team uses Pest with test impact analysis, and developers work in git worktrees (or frequently switch branches), so cache keying by absolute path is a known friction point.
• You already run, or want to run, a preflight/pre-merge gate and want it enforced consistently rather than reimplemented per repo.
• You are paying real money or wall-clock time for redundant full test runs on the first and last commit of each branch.
• You maintain in-house scripts that copy Pest caches or wrap composer test / PHPStan / retries — the maintenance cost is a direct argument for consolidating.
• You need baseline inheritance on base branches other than main.

Look elsewhere or wait when:

• You don't use Pest's TIA (the core value is inert without it), or you're on PHPUnit-only stacks.
• Your workflow is single-checkout, single-branch, with fast suites — the overhead likely outweighs the gain.
• You require a stable, 1.x-or-later API. This is v0.4.0; commands, flags, and behaviors are still moving. The release itself changes enforcement (preflight now refuses certain flags) — that's a behavior change that can break existing CI pipelines overnight.
• You need vendor-backed SLAs, long-term support guarantees, or an audit trail — this reads as an opinionated internal-style toolkit, not an enterprise product.
• Lock-in risk: adopting toolkit:tia means your baseline lifecycle is coupled to a pre-1.0 package that is intentionally tracking Pest's internal cache format. That format tracking is a feature today and a fragility tomorrow.

How to Pitch It (Stakeholders)

For executives (one paragraph): Our Laravel teams currently burn a full test-suite run at the start and end of every ticket branch, because the test-impact-analysis cache doesn't travel between git worktrees and branches. This release adds a small command that lets a merged branch publish its test baseline and lets new checkouts inherit it — so routine changes run only the tests they touch, while keeping the final gate unfiltered so we don't trade speed for coverage. It also closes a hole where a filtered preflight quietly disabled test-impact analysis and re-ran everything, and removes internal scripts teams were maintaining to work around it. Net effect: faster feedback per ticket and lower CI spend, with the same or stricter quality gate. Caveat: this is a pre-1.0 tool, so we should pilot it on one or two repos before standardizing.

For engineering: toolkit:tia (check / inherit / publish) replaces the per-repo cache-copying scripts; it understands Pest's cache format and works on any base branch, not just main. toolkit:preflight now hard-fails on coverage gates that pass selection flags like --exclude-testsuite=Browser, since those disabled TIA and recorded nothing to publish — the "Running Tests" guideline documents the rationale. Two fixes land alongside: the guard hook no longer blocks pest --baseline, and two PHPStan errors are resolved. toolkit:retry gains --no-notify for parity with sibling commands. Expect to update CI configs if any pipeline currently filters the preflight suite; that will now fail loudly by design.