- How do I install Laravel Tackle?
- Install Laravel Tackle with Composer using composer require jordandalton/laravel-tackle. Because it is a developer-facing Artisan CLI, the recommended placement is require-dev so it stays out of production dependency graphs and deploy artifacts. After installing, publish the package config and supply your model provider credentials via local environment variables before running the ai:code command in a scratch repository.
- Should I install Laravel Tackle as a require-dev or require dependency?
- Use require-dev. Tackle ships an Artisan surface (ai:code) intended for developer workflows such as reading code, running tools, and running tests, not application runtime behavior. Keeping it in require-dev avoids shipping the tool and its dependencies to production.
- Which Laravel and PHP versions does Laravel Tackle support?
- The provided overview and README excerpt do not list a supported Laravel or PHP version matrix. Check the package composer.json and release notes against your baseline before rollout. Confirm support for the oldest Laravel version you run in production and the PHP version in your CI base image.
- What does the ai:code command actually do?
- Tackle provides an Artisan ai:code interactive session that lets AI agents operate inside your Laravel application. It uses Git-aware context, can read code, execute tools, and run tests, and enforces safety boundaries at the framework level. It also supports webhook-triggered tasks, according to the technical evaluation.
- How are AI provider credentials handled in Laravel Tackle?
- Provider credentials are supplied through the package's publishable configuration, typically via environment variables. Keep those keys in local env or a secrets manager and never commit them. The provided notes do not document telemetry or transcript retention, so review those areas with security and legal before broader rollout.
- Does Laravel Tackle work in CI or non-interactive, headless environments?
- The evaluation calls out non-interactive behavior (CI, headless, non-TTY shells) as a gating factor for mainstream adoption, so test it headless in a sandbox first. The documented behavior when stdout is not a TTY is not stated in the provided material. Run it in your containerized CI shell and confirm it degrades gracefully rather than failing the job.
- What safety boundaries does Laravel Tackle enforce?
- Tackle is described as enforcing safety boundaries at the framework level, with the runtime layer designed so agents can act without unrestricted access. The provided overview does not enumerate the specific boundary controls or configuration flags. Review the published config and package docs to confirm which tools and actions are permitted in your environment.
- How does Laravel Tackle read Git state, such as staged versus unstaged changes?
- Tackle builds Git-aware context and, per the evaluation, quiets Git summaries unless the tracked diff actually changes. The provided notes do not document exactly how it reads working tree versus index or staged versus unstaged changes. Validate behavior in shallow clones, worktrees, and monorepos before relying on it in those layouts.
- Can I use Laravel Tackle session transcripts as an audit trail?
- A recent patch improved transcript fidelity by stopping response chunks from being redrawn repeatedly and only showing Git summaries when the diff changes. That helps if you want to treat session logs as an audit trail, but transcripts are still developer tooling output. Confirm retention, storage, and provider data handling with security before using them as an audit record.
- How actively is Laravel Tackle maintained, and what is the release cadence?
- The technical evaluation notes a high release cadence, with v1.56.x implying a long, frequent patch tail. Pin the package version and set an upgrade cadence; the risk is contained to dev tooling because the package is installed as require-dev. Also confirm the maintainer model and mean time to patch for rendering or provider-SDK regressions.