AutoPipeline_2025

What it does

AutoPipeline_2025 is a C# plugin for Cimatron 2025 that turns the repetitive setup of a two-sided 5-axis CNC programming environment into a guided, repeatable workflow.

Starting from a prepared CAD part, AutoPipeline coordinates the sequence required to create the NC programming environment. It validates required coordinate systems, collects the SETUP-1 and SETUP-2 machining templates, analyzes part geometry to recommend an appropriate fixture, creates the NC environment, guides both fixture-placement stages, restores protected fixture state, and presents a clear completion state.

The goal is guided automation rather than one-click manufacturing. Predictable and verifiable operations are automated, while the programmer remains responsible for manufacturing decisions and native Cimatron interactions that require human judgment.

Why I built it

Preparing a two-sided CNC job requires several pieces of state to remain aligned: setup and pickup coordinate systems, machining templates for both sides, fixture selection and placement, active Cimatron sets, and native dialogs that must be completed in the correct order. When these operations are handled independently, the programmer must remember and coordinate the entire sequence while moving between the Part, NC document, templates, fixtures, and native tools.

AutoPipeline makes that sequence visible and repeatable. It reduces repetitive setup work and helps prevent missed steps without removing the CNC programmer from decisions that matter.

Build Week work and AI-assisted development

The project foundation existed before OpenAI Build Week. During Build Week, Codex was used extensively to extend and refine AutoPipeline_2025.

The primary Codex implementation session was 019f66f4-99db-7ac3-8ca6-c9106a2bb8e4. Local Codex records identify that primary implementation thread as using GPT-5.5. It contains the majority of the core Build Week implementation work, including the guided pipeline wizard, fixture recommendation integration and reuse, single-instance workflow and dialog lifecycle management, finalization improvements, monitor-aware window placement, removal of obsolete workflow pieces, and the final completion state.

A separate related Build Week Codex session, 019f7220-8263-7300-b2d8-62840492eda1, is locally verified as using GPT-5.6 Sol. GPT-5.6 Sol was used for project auditing, Build Week evidence and documentation, repository preparation, diagnostics, and troubleshooting of the AutoPipeline workflow, including diagnostic logging work and development of a separate read-only procedure-state probe.

Codex accelerated the development workflow by helping inspect and reason about the existing C# codebase, implement and refine changes, build and validate the project, investigate Cimatron API behavior, and document evidence of the Build Week work. The manufacturing decisions that define the workflow—what should be automated, where programmer judgment is required, how fixture selection should work, and what state must be protected—were based on real CNC programming experience.

GPT-5.6 Sol is not an embedded runtime dependency of the plugin, and the project is not represented as having been entirely built with GPT-5.6.

Testing

AutoPipeline_2025 is a specialized plugin that runs inside a licensed Cimatron 2025 environment on Windows. A complete runtime test requires a compatible Cimatron installation, authorized Cimatron API/interop assemblies, the required NC and machining templates, fixture assets, and a prepared Cimatron Part containing the coordinate systems expected by the workflow.

The repository intentionally excludes licensed Cimatron binaries, proprietary manufacturing templates, fixture models, and customer CAD/CAM data. The public demo video is the primary way for judges without access to the licensed Cimatron environment to see the complete workflow running. Static source review can verify the pipeline orchestration, wizard implementation, validation, fixture recommendation logic, single-instance protection, native command integration, protected-state restoration, and completion behavior.

In April, neither did I didn’t know what an API was nor did I have much experience with AI. Today, I’m submitting a working C# plugin that can make real changes to save money in an industry looking for it. OpenAI has helped me learn how to use all of it.

Developed by Aaron Beahm with the support of Apollo Tool Inc.

Built With

• .net-framework-4.8
• c#
• cimatron-2025-api
• codex
• gpt-5.6-sol
• winforms