MendixToolkit

Spin up an HTTP/SSE wrapper around a stdio MCP server so Maia (Studio Pro 11.8+ MCP Client) can register it. Maia accepts only HTTP-Streamable or SSE — NOT stdio — so any MCP registered with Claude Code via `claude mcp add` (which is mostly stdio) is invisible to Maia by default. This skill emits a one-process HTTP bridge plus the exact Maia-Chat configuration steps Neo pastes into Studio Pro. Auto-invoke when Neo says "make X available in Maia", "wire MCP X to Maia", "register Y in Studio Pro AI", or types `/maia-mcp-bridge <mcp-name>`. Read the Maia MCP reference before running: https://docs.mendix.com/refguide/maia-mcp/

GATED WRITE. Clone an existing well-built Mendix page (ideally a gold page) to a new name with optional entity rebinding, via the mxcli MDL round-trip. Routes through the shadow-write engine (7 safety pillars; live .mpr untouched until atomic swap). Pre-classifies the source page as SIMPLE / COMPLEX / UNCLONEABLE based on parser-incompatible patterns observed 2026-04-27. Refuses UNCLONEABLE; requires --allow-lossy on COMPLEX. Auto-emits a /mx-handoff sheet for any widgets stripped during the clone. Invocable as /mx-page-clone-from-template <Source.Page> <Target.Page>.

Gracefully close Mendix Studio Pro and its child processes (running runtime, MxBuild, Java helpers) so subsequent CLI ops can run without tripping the Studio-Pro-lock check. Also cleans up Mendix toolchain orphans left over from a Studio Pro crash (node/esbuild/sass/mxbuild under \Mendix\<ver>\(modeler|runtime)\ whose parent studiopro is dead) so Studio Pro can relaunch. Defaults to graceful (sends WM_CLOSE, lets Studio Pro prompt about unsaved changes). Auto-invoke when a /mx-shadow-write or /mx-scaffold-microflow needs Studio Pro closed and Neo has confirmed it's safe to do so, OR when Studio Pro is failing to launch and orphan toolchain processes are suspected. Also invocable as /close-studio-pro [project-name].

Drive a Playwright browser test against a page on the running TestOSApp3 testbed (port 8082). Auto-asserts the testbed fingerprints (URL :8082 + "Test Project" ribbon) before any further assertion -- this is a HARD safety gate against accidentally driving the live OneSource project. Uses the Playwright MCP server (auto-attached to Claude Code as `playwright`). Captures screenshot + accessibility snapshot. Auto-invoke after a /mx-handoff completes, after any page change, and as part of any verification flow that needs to prove the rendered UI works. Also invocable as /mx-test-page [page-path-or-name].

GATED WRITE. Add a new entity (or attribute on an existing entity, or association) to a Mendix project by wrapping mxcli's create commands with the 8 safety pillars. Hard precondition: a fresh /mx-discover inventory must exist (≤24h old) so we can verify the new name doesn't collide. Pre-tag, clean-tree, Studio Pro lock, version, pre/post mx check, auto-restore on fail. Migration risk is flagged before the write. Invocable as /mx-entity-add.

GATED WRITE. Scaffold a new Mendix microflow from a structured spec by wrapping mxcli's create microflow command with the 8 safety pillars from rules/mendix-safety.md. Pre-tag the project, verify clean tree + Studio Pro closed + Mendix version, run mx check pre/post, auto-restore on failure. Refuses to proceed unless every pillar passes. Invocable as /mx-scaffold-microflow <spec-or-description>.

Resolve the Mendix Studio Pro version a project was last opened with, and detect mismatches against installed Studio Pro / mx CLI / mxcli binaries on this machine. Read-only. Auto-invoke at the start of any Mendix session in a project, before running mx convert, and when a build/check produces unexplained errors. Also invocable as /mx-version.

Mid-session orchestrator. Given the current project state and what was just done, propose the next move. Lighter than /start-here -- skips the health check, focuses on "what's the highest-leverage next thing." Spawns mendix-orchestrator agent for the synthesis. Use when a skill just finished and the user asks "what next?" or "ok, now what?".

First-message orchestrator for any session opened in MendixToolkit or its testbed. Auto-runs a health check (MCP servers, scheduled tasks, ports, mx version), reads MISSION.md + LEARNINGS.md + recent git activity to recap the project state, then proposes the top 3 next steps with concrete commands and rationale. Use when the user opens a fresh session and asks "what's next?" / "go" / "where did we leave off?" / anything ambiguous on the first message. Also invocable as /start-here.

Invoke a named recipe from WORKFLOWS.md by name. Each workflow is a defined skill chain with a verification step and rollback path. Use when the user types /workflow <name> or asks "run the X workflow" -- the skill reads WORKFLOWS.md, finds the matching recipe, and executes the steps in order with proper preconditions checked at each step. Pauses for explicit go-ahead before any gated-write step.

Fresh-context reviewer for a proposed Mendix design. Reads the /mx-discover inventory and the proposed design, then audits independently of the original session for naming collisions, anti-patterns, missing test coverage, security concerns, marketplace-overlap (use vs. build), and migration risk. Read-only against the model. Spawn before committing to any non-trivial Mendix design. Also invocable as /mx-design-review.

Mendix-specialized domain inventory. Use BEFORE designing any Mendix feature, redesigning a microflow, adding entities, or extending an existing app. Auto-invokes the mendix-expert subagent, lists in-scope entities + associations + relevant microflows + available marketplace modules, and produces a DOMAIN_INVENTORY block. Refuses to proceed to design until the inventory is on the table. Born from the Apr-2026 Next_Best_Action rewrite.

Mendix project health check. Read-only. Wraps `mx check` (Mendix's bundled validator) plus mxcli's lint/report passes. Surfaces errors, warnings, deprecations, performance concerns, security flags, and convention violations -- triaged by severity. Auto-invoke before any commit / push of a Mendix project, after pulling someone else's changes, and after any auto-conversion (mx convert). Also invocable as /mx-doctor.

Convert a plain-English Mendix feature description into a structured spec markdown file referencing inventory items by exact name. Output is a spec, not a model change. Auto-invoke when the user describes a Mendix feature in natural language and is about to start designing it. Also invocable as /mx-feature-spec <description>.

Generate copy-paste-ready Studio Pro instruction sheets for the rare moments when only a human can perform the action (drag a widget, set a CSS class, configure a layout). Reads the project's theme catalog and widget conventions to produce EXACT instructions -- specific paths, classes, widgets, bindings -- not vague guidance. Auto-invoke when an automated path is blocked and a human-in-loop step is required. Also invocable as /mx-handoff <task-description>.

Blast-radius analysis for any Mendix entity / attribute / microflow / page. Read-only. Answers "what breaks if I rename / delete / change X?" before you change anything. Auto-invoke before any rename, delete, or signature change in a Mendix project; also invocable as /mx-impact <target>. Wraps mxcli's callers, callees, refs, and impact subcommands. Zero risk -- never writes.

Read-only audit of a Mendix page against project doctrine. Flags SmartHub class usage (DEPRECATED), classes not in the OS2 catalog (snowflakes), missing design-properties bindings, page-template deviations (Template A or B), accessibility-tree red flags, and any class that does not exist in the project's theme catalog. Uses the mendix-context MCP tools (theme_catalog, os2_class_groups, validate_classes, theme_architecture) to ground every finding in doctrine. Auto-invoke before merging any page change, after /mx-handoff completes, and as part of the standard pre-codex flow for Mendix work. Also invocable as /mx-page-audit <page-name-or-path>.

Fast structured search across a local Mendix project. Read-only. Wraps mxcli's search and SQL-like select to answer questions like "find every microflow that references entity Customer", "which pages bind to entity X", "what microflows have empty error handlers". Auto-invoke when the user asks a model-spanning question that grep on .mpr (binary) cannot answer. Also invocable as /mx-search <query>.