migo-ovo/plc-skill
Developer toolsこのページは AI により自動翻訳されています。技術的な詳細は原文をご確認ください。 原文を見る
A PLC-focused skill repository that streamlines industrial control workflows, troubleshooting, and automation-related tasks.
概要
Production-grade AI agent skill for industrial PLC programming. It combines a vendor-neutral engineering layer with vendor-aware routing so agents can produce structured, reviewable IEC 61131-3 control logic instead of generic PLC pseudocode. PLC Skill is an extensible knowledge base for OpenClaw, Cursor, Claude Code, and other LLM-powered development tools that need to reason about PLC logic, control sequences, interlocks, alarms, and vendor-specific automation environments. The skill is organized as a two-layer system: 1. : Cross-vendor engineering guidance for scan-cycle reasoning, state machines, output ownership, interlocks, debugging, code review, and IEC 61131-3 style logic. 2. : Platform-specific routing for terminology, syntax, memory models, project structure, software behavior, and official documentation. This separation helps an AI assistant avoid mixing vendor ecosystems, making up unsupported syntax, or flattening every PLC request into vague IEC-style code.
README
以下は原文 README の抜粋です。原文を見る。
PLC Skill
Production-grade AI agent skill for industrial PLC programming. It combines a vendor-neutral engineering layer with vendor-aware routing so agents can produce structured, reviewable IEC 61131-3 control logic instead of generic PLC pseudocode.
Overview
PLC Skill is an extensible knowledge base for OpenClaw, Cursor, Claude Code, and other LLM-powered development tools that need to reason about PLC logic, control sequences, interlocks, alarms, and vendor-specific automation environments.
The skill is organized as a two-layer system:
- Common PLC layer: Cross-vendor engineering guidance for scan-cycle reasoning, state machines, output ownership, interlocks, debugging, code review, and IEC 61131-3 style logic.
- Vendor layer: Platform-specific routing for terminology, syntax, memory models, project structure, software behavior, and official documentation.
This separation helps an AI assistant avoid mixing vendor ecosystems, making up unsupported syntax, or flattening every PLC request into vague IEC-style code.
Capabilities
- Logic design and generation for Structured Text (ST), Ladder Diagram (LD), Function Block Diagram (FBD), and Sequential Function Chart (SFC) oriented work.
- State machines and sequencers with explicit state ownership, transition conditions, and output mapping.
- Code review and refactoring for scan-cycle risks, multiple writers, hidden latches, unclear interlocks, and maintainability problems.
- Debugging and troubleshooting using observable-first workflows for timers, edge triggers, resets, alarms, and sequence faults.
- Vendor-aware routing for cues such as TIA Portal, Studio 5000, GX Works, Sysmac Studio, TwinCAT, CODESYS, and related platform terminology.
See SHOWCASE.md for examples of how the skill changes typical AI-generated PLC output.
Support Depth
| Depth | Vendors and environments | Expected behavior |
|---|---|---|
| Mature path | Mitsubishi FX3U + GX Works2 + Structured Project + ST | Deepest bundled guidance, examples, and review patterns. |
| Targeted modules | Siemens, Rockwell / Allen-Bradley, Omron | Vendor routing, rules, examples, and focused eval coverage. |
| Baseline modules | Beckhoff, Schneider, CODESYS, Delta, Keyence, Panasonic | Recognition, reference routing, cheatsheets, and official-doc indexes. |
Do not assume every vendor path has the same depth. When a platform is unknown or only partially covered, the skill should answer from the common PLC layer and mark vendor-dependent details clearly.
When to Use It
推奨ツール
別のキーワードを試すか、フィルタを外してください。
インストール
npx skillfish add migo-ovo/plc-skill