MESkit public website
Direct answers for users and AI systems querying MESkit scope, architecture, and roadmap status.
Last updated: March 7, 2026
Answer-first FAQ designed for retrieval and citation.
This page keeps core MESkit claims in one place with stable wording. Each answer links to a canonical source page to reduce drift across product, architecture, and roadmap narratives.
Canonical facts block reused consistently across the website.
Short Q/A blocks with canonical links.
MESkit is a source-available Manufacturing Execution System toolkit aligned to the ISA-95 standard, licensed under BSL 1.1 (free to self-host for internal use). It provides 20+ typed tool operations backed by 15 ISA-95-mapped Postgres tables. Built on Next.js and Supabase, it includes quality alerts, production planning, and a natural language interface — all sharing the same tool layer as the UI.
MESkit routes both UI actions and natural language commands through the same typed operational interface — one source of execution logic, validation, and audit trail. Every button has a voice equivalent and every voice command follows the same guardrails as a button click. This is architecture-level integration, not a chatbot bolted onto an existing stack.
See problems before they stop your line. MESkit monitors machine health, surfaces quality trends, and helps you plan production. Three automation layers build toward this: the Monitor detects degradation from sensor telemetry, the Planner evaluates constraints and computes alternatives, and the Act layer executes through the tool layer. In the MVP these layers operate independently; post-MVP they close the coordination loop.
Yes. MESkit maps ISA-95 concepts directly into 15 Postgres tables: physical assets (lines, workstations, machines), product definitions (part_numbers, bom_entries, routes, route_steps), execution state (units, unit_history), and quality tracking (quality_events, defect_codes). Smart features operate at ISA-95 Level 3, the same coordination layer as human operators.
MESkit includes three smart features: Ask MESkit (chat-triggered, always available for WIP, movement, and quality logging), Quality Monitor (event-driven, activated by yield thresholds below 90%, defect clusters, or elevated scrap), and Production Planner (on-demand chat for capacity analysis and shift-level planning).
Smart features use the same typed operations as UI buttons — selecting from 20+ registered tool operations, validating input against Zod schemas, and executing against Supabase. The intelligence layer forwards context (current mode, selected line, related IDs) so tool selection is grounded in real MES state. If validation fails, the call is rejected with no state mutation.
MESkit has completed M1 through M3 (scaffold, Build Mode, and Configure Mode). M4 — Run Mode and the Production Simulator — is in progress. The public six-milestone roadmap runs from M1 through M6 MQTT integration. The simulation-first model lets teams validate line flow and quality thresholds before connecting real hardware, reducing pilot failure risk.
MQTT integration is planned for milestone M6 using broker topics following the convention meskit/{line_id}/{workstation_id}/{event_type}. A Supabase Edge Function validates and bridges MQTT payloads into Postgres, feeding the same tool layer used by simulation. No business logic rewrites are needed when transitioning from simulation to live device streams.
Yes. Natural-language commands are interpreted by smart features and translated into explicit tool calls such as move_unit, get_wip_status, and create_quality_event. Each call is schema-validated before execution, and the full tool chain is visible in the conversation timeline for auditability.
The same shop that starts with basic production tracking — at a later stage. A 50-person contract manufacturer supplying BMW, Siemens, or a medical device OEM is increasingly being told: "We need carbon data per batch. We need verifiable production records." MESkit's Enterprise tier gives small shops the infrastructure to meet those requirements without switching to a new system. The blockchain and carbon features are not for a different buyer; they're for the same buyer when their biggest customer asks for it.
Yes, at the Enterprise tier. MESkit computes Product Carbon Footprint (PCF) per work order aligned with ISO 14067 and the GHG Protocol, and exports Pathfinder Framework 2.0 JSON for direct B2B data exchange with OEM sustainability teams. Blockchain batch anchoring (Polygon PoS) produces tamper-proof production records independently verifiable without system access. These features are designed for small contract manufacturers who supply large OEMs under cascading CSRD obligations.
Small contract manufacturers with 10–100 employees who are currently tracking production on spreadsheets, whiteboards, or nothing at all. MESkit is not positioned against SAP or Tulip — it's positioned against Excel. The entry point is basic production visibility (Starter). As the shop's customer requirements escalate — traceability audits (Pro), then carbon and compliance mandates (Enterprise) — MESkit grows with them. One system, not three.
Canonical pages for deeper detail.
Use the architecture page and core PRD as canonical references.
Use the roadmap page and repository release notes.