Mendix-aware Claude Code overlay: agent army v1 (mendix-doer + 4 reviewers + researcher), mendix-context + mendix-inventory MCPs (40 tools), 7-pillar shadow-write engine, Studio Pro 11.8+ / Maia / mxcli integration. Built atop NewClaude.
This repository also configures its own agents. See what MendixToolkit tells them →
Command Claude Code
Gracefully close Mendix Studio Pro + its children (runtime, MxBuild, Java helpers) and clean up orphan toolchain from a previous crash. Sends WMCLOSE first; reports counts of what was closed. Default project: TestOSApp3 testbed; default port: 8082.
Command Claude Code
Fresh-context reviewer for a proposed Mendix design. Audits against the /mx-discover inventory for naming collisions, anti-patterns, missing tests, security, marketplace overlap, migration risk. Read-only.
Command Claude Code
Mendix-specialized domain inventory. Auto-uses mendix-expert subagent. Refuses to proceed to design until in-scope entities, microflows, and module overlaps are on the table. Born from the NextBestAction rewrite.
Command Claude Code
Mendix project health check. Wraps mx check (authoritative) + mxcli lint/report (advisory). Reports errors, warnings, deprecations, perf concerns, security flags. Read-only.
Command Claude Code
GATED WRITE. Add a Mendix entity / attribute / association with the 8 safety pillars + collision check + design-review gate + migration-risk flag.
Command Claude Code
Convert a plain-English Mendix feature description into a structured spec referencing inventory items by exact name. Spec output, not model changes.
Command Claude Code
Generate copy-paste-ready Studio Pro instruction sheets when a human-in-loop step is required. Exact paths, classes, widgets, bindings -- no vague guidance. Auto-runs verification (mx-doctor + impact + playwright) after the human replies "done".
Command Claude Code
Blast-radius analysis for a Mendix entity/attribute/microflow/page. Read-only. Answers "what breaks if I rename or change X?" before the change. Wraps mxcli refs/callers/impact.
Command Claude Code
Read-only audit of a Mendix page against OS2 doctrine. Flags SmartHub usage (BLOCK), hallucinated classes (BLOCK), snowflake classes (FLAG), template deviations, hardcoded-where-design-property exists. Cites doctrine sources for every finding.
Command Claude Code
GATED WRITE. Clone a Mendix page to a new name + optional entity rebinding via the mxcli MDL round-trip. Routes through the shadow-write engine (7 safety pillars). Pre-classifies the source as SIMPLE / COMPLEX / UNCLONEABLE; refuses UNCLONEABLE, requires --allow-lossy on COMPLEX. Auto-emits a /mx-handoff sheet for…
Command Claude Code
GATED WRITE. Scaffold a new Mendix microflow with the 8 safety pillars. Pre-tag, clean-tree check, Studio Pro lock check, version check, pre/post mx check, auto-restore on fail.
Command Claude Code
Fast structured search across a local Mendix project. Wraps mxcli search and select. Read-only. Answers model-spanning questions grep on .mpr cannot.
Command Claude Code
Drive Playwright against the testbed (port 8082). MANDATORY testbed-fingerprint guard (URL :8082 + "Test Project" ribbon) before any other assertion -- safety gate against driving the live project.
Command Claude Code
Resolve the Mendix Studio Pro version of a project, inventory installed Studio Pro / mx / mxcli binaries, and report mismatches. Read-only.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: