Product Decisions This Supports
- Editorial quality in the admin. The package adds a symbol row to the Filament rich text editor so authors can insert typographic characters (dashes, quotes, spaces, etc.) without pasting from Word or a character-map website. That directly supports copy standards and reduces formatting drift in CMS-driven content.
- Build vs. buy on a small UI surface. This is a narrow, low-blast-radius capability sitting on top of a platform component we already own. Buying it (free, MIT-style plugin) is cheaper than maintaining our own editor toolbar extension across Filament major upgrades.
- A cheap, low-risk roadmap item. It fits as a quality-of-life improvement in a minor release for content-editing teams — not a differentiator, not something to build a launch around.
- Scoping decision: adopt narrowly or not at all. The feature is only valuable to teams whose editors actually type special characters (editorial, marketing, localization/translation workflows). For internal CRUD forms, the value is near zero and the dependency isn't worth it.
- Risk posture is now a decision input. v0.1.1 shipped with its assets never loading, meaning the package's headline feature was invisible to every user for a release. That tells us the maintainer's release process doesn't include a visual smoke test. This doesn't disqualify the package, but it changes the adoption plan from "install and forget" to "pilot and verify."
When to Consider This Package
Adopt (pilot) when:
- You are on Filament with the rich text editor in active editorial use, and authors have asked for special-character input.
- You can pin an exact version and absorb pre-1.0 churn — this is 0.1.x, so breaking changes to the modifier API and rendering are expected.
- You have someone who can visually verify the editor in both light and dark mode on staging before each bump. The two fixes in this release (assets not loading, transparent fullscreen in light mode) are exactly the class of bug that unit tests miss and eyeballs catch.
- You previously added your own render hooks to load the package assets as a workaround for v0.1.1 — remove those now; the package loads its own script and stylesheet, scoped to pages that call
typebar().
Look elsewhere or wait when:
- You need a supported, versioned dependency with an SLA, long-term maintenance guarantees, or a multi-maintainer bus factor. This is a single-author package at 0.1.x.
- You need a full editor replacement or markdown authoring — this is a toolbar-level enhancement, not an editor.
- Your team cannot run manual UI checks per release, or you auto-update dependencies without review.
- Your editors don't need symbol input; adding a plugin for an unused symbol row is pure surface area.
Also note for engineering: documentation has moved out of the repo to docs.aw.codes, and the README is now a landing page. Internal runbooks should link there rather than to GitHub, otherwise onboarding docs will rot.
How to Pitch It (Stakeholders)
For executives: This is a small, free plugin that makes the CMS editor better for the people writing customer-facing copy — fewer pasted-in formatting artifacts, more consistent typography. Cost is a few hours of integration and a staging check per upgrade, not a platform commitment. It is pre-1.0 and single-maintainer, so we're piloting it in one admin panel rather than standardizing on it; if it stalls, removal is a one-line change.
For engineering: A Filament plugin exposing typebar() with chainable modifiers. This release fixes two user-visible defects — assets are now registered and loaded only on pages that need them, and the fullscreen editor's light-mode styling uses a defined CSS variable again. CI has been normalized to a shared workflow pinned by commit SHA, with phpstan and rector paths corrected, so static analysis is trustworthy. The caveat: 0.1.1 shipped with no assets loading at all, so treat visual regression as a real risk — pin the version, and add a smoke check on the editor in light and dark mode to your upgrade checklist. Docs now live at docs.aw.codes.