
Process-flow workspace where people and AI build nano devices together
Arpelier is a process-flow workspace for semiconductor and nano devices, made for people and AI to work on together. Each step of a process flow (deposition, lithography, etch, implant, CMP and more) becomes a 3D structure you can inspect.
https://arpelier.com/mcpArpelier builds 3D semiconductor device structures from process flows (a geometric model with lengths in nm; not a fab recipe or an electrical sign-off). This server needs no account. It validates and computes Flow 1.0 JSON documents (https://arpelier.com/schema/flow-1.0.json) with the same engine as the Arpelier web app. Eight of its nine tools are read-only; keeps a Flow as a link record for 7 days. It never writes to anyone's Arpelier projects. To continue calculations across calls it keeps computed states for up to an hour (see the hosted limits below).
Tools:
returns Arpelier's published reference documentation: the Flow format, process kinds and fields (processes.md), material IDs (materials.md), mask polarity (masks.md), example Flows and goal-specific task pages. With no path it returns the index (/llms.txt).
checks a Flow with the web app's own parser and lists issues with JSON paths.
computes up to a step and returns per-material volumes, the top-surface height range and a summary of voids (empty space below the surface).
returns a section: a character map and, for vertical sections, the exact layer stacks (material, fromNm, toNm) read from the engine.
with image true also returns a PNG of the section colored by material.
returns a PNG 3D view of the computed structure from one corner of the domain; hiddenMaterialIds leaves materials out to show what is inside or behind them.
None in the store listings or on the server