Statamic Marketing Toolkit Laravel Package

Product Decisions This Supports

• Whether to invest in AI/answer-engine discoverability. The package produces llms.txt artifacts for a Statamic site — a curated, machine-readable map of content aimed at LLM crawlers and agents. That makes "GEO / AI search visibility" a channel you can test cheaply rather than a strategy you have to fund speculatively.
• Build vs. buy for generator tooling. The alternative is 100–300 lines of in-house code that walks your content tree, applies ordering/filtering rules, and writes a static file. Adopting this addon trades that maintenance for an external dependency — a trade worth making only if the output format is something you'd otherwise hand-roll.
• Whether generated artifacts are committed or built at deploy time. v0.29.1 fixes llms.txt retaining its order between builds, which is the prerequisite for treating the file as a reviewable artifact: stable diffs in pull requests, deterministic CI output, no spurious "changed file" noise on every deploy. This single fix is what lets a team decide "commit it and code-review changes" instead of "generate and forget."
• Content-ops process. Once output is deterministic, "ship an updated llms.txt with every major content launch" becomes a checklist item rather than an ad-hoc task, and marketing can own the config without engineering involvement.
• Dependency governance. The project is pre-1.0 and iterating quickly. That supports a decision to pin exact versions and review upgrades deliberately, rather than floating the constraint.

When to Consider This Package

Consider it when:

• You run a Statamic site (Laravel-based, flat-file CMS) and want llms.txt and similar marketing artifacts without custom code.
• Marketing owns content and wants configuration in versioned files, with engineering not maintaining bespoke PHP for each new artifact.
• You value reproducible builds — the ordering fix in v0.29.1 means you can diff generated output in CI or commit it to the repo without churn.
• You're comfortable running a small, single-maintainer, pre-1.0 open-source addon and pinning it.

Look elsewhere when:

• You're not on Statamic (WordPress, Drupal, a JS framework, or a hand-rolled Laravel app) — the value here is CMS integration, not the file format.
• You need a supported vendor contract, SLAs, or a guaranteed upgrade path. At 0.x, minor releases may carry breaking changes even when a given patch says "nothing to do."
• You need a full, mature SEO suite with per-locale guarantees, redirect management, and analytics tie-ins — verify current feature coverage before assuming this covers it.
• You can't accept the format risk: llms.txt is a community proposal, not a ratified standard, and crawler adoption is unproven. This should be a low-cost, reversible bet.

How to Pitch It (Stakeholders)

To executives: This is a cheap, reversible experiment in AI search visibility. We generate a machine-readable index of our content aimed at AI assistants and crawlers, using an existing open-source addon rather than custom development — no licensing cost, a few hours of setup, and we can remove it by deleting one config entry. Framing matters: AI referral traffic is not yet a proven channel, so treat this as a small hedge we can measure, not a commitment. Success criteria should be defined up front (AI-referral sessions, crawler hits on the file) with a decision point to expand or drop.

To engineering: It's a config-driven Statamic addon — no core modifications, no runtime coupling to our application code, and output is static. v0.29.1 specifically fixes ordering instability in llms.txt between builds, which means generated output is now deterministic: clean diffs, no false-positive change detection in CI, and safe to commit as a reviewable artifact. Caveats to flag honestly: the package is pre-1.0 with a single maintainer, so we pin the version, review upgrade notes rather than auto-bumping, and keep the generated file in our own repo so removing the dependency is trivial. Our assessment here is based on release notes only — a short spike to read the source and confirm behavior on our content model should precede adoption.

To marketing: You get an llms.txt you can steer without filing engineering tickets, and because output is now stable, changes you make show up as clear, reviewable diffs.