KUKA KRL MCP Server
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@KUKA KRL MCP Server Decode KSS error 1376 and give me the SmartPAD reset steps"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
KUKA KRL MCP Server
Official open-source Model Context Protocol (MCP) server for KUKA Robot Language (KRL) and KUKA System Software (KSS 8.3–8.7).
Empowers AI coding assistants (Cursor, Google Antigravity, Claude Desktop, Claude Code CLI, Windsurf, Copilot) with industrial robotics knowledge, motion safety analysis, KSS error diagnosis, workspace continuous verification, and live controller telemetry.
Architecture: Hybrid Three-Tier Pipeline
flowchart TD
subgraph Clients["AI Clients & IDEs"]
Cursor["Cursor AI"]
Antigravity["Antigravity IDE"]
Claude["Claude Desktop / CLI"]
Windsurf["Windsurf"]
end
subgraph MCP["kuka-krl-mcp (stdio)"]
T0["Tier 0: Standalone Knowledge<br/>(System Vars, KSS Alarms, Syntax, Offline Lint)"]
T1["Tier 1: Workspace Intelligence<br/>(SRC-DAT sync, Project diagnostics)"]
T2["Tier 2: Pro Industrial Suite<br/>(Live Telemetry, Auto-Repair, Control Flow)"]
end
subgraph IDE["LiskinLabs KUKA KRL Extension"]
Bridge["Local Bridge Server<br/>(127.0.0.1 + Token)"]
KRC["KUKA Robot Controller<br/>(KRC4 / KRC5 / Virtual)"]
end
Clients --> MCP
T0 -->|"Offline / Free"| MCP
T1 --> Bridge
T2 --> Bridge
Bridge --> KRCTier 0: Standalone Knowledge (Free & Offline):
Instant access to KSS system variables (
$TOOL,$BASE,$OV_PRO,$VEL,$ACC,$IN,$OUT,$TIMER).KSS alarm catalog (codes 1, 14, 20, 67, 71, 1376, 1422, 2858) with root cause analysis and SmartPAD reset steps in 6 languages.
KRL motion syntax reference (PTP, LIN, CIRC, SPLINE, INTERRUPT, TRIGGER, FOLD).
Fast static linter for code snippets.
Tier 1: Workspace Intelligence (Requires KUKA KRL Extension):
Project-wide continuous diagnostics, SRC↔DAT pair integrity, CIRC 2-point geometry, physical I/O ranges.
Tier 2: Professional Industrial Suite (Requires KUKA KRL Pro):
Deep industrial safety audit (advance run hazards, turn unwinding, payload limits).
Safe auto-repair with verified code diffs.
Real-time controller telemetry reading (actual TCP position, joint angles, active program line pointer, SmartPAD alarms).
Related MCP server: Studio5000 AI-Powered PLC Programming Assistant MCP Server
Quickstart & Client Configuration
1. Cursor
Add to your project's .cursor/mcp.json or Cursor Settings > MCP:
{
"mcpServers": {
"kuka-krl": {
"command": "npx",
"args": ["-y", "kuka-krl-mcp"]
}
}
}Note: If you have the LiskinLabs KUKA KRL Extension installed in Cursor, the extension automatically registers the MCP server with the local bridge credentials at startup!
2. Google Antigravity IDE
In Antigravity settings or ~/.gemini/antigravity/mcp_config.json:
{
"mcpServers": {
"kuka-krl": {
"command": "npx",
"args": ["-y", "kuka-krl-mcp"]
}
}
}3. Claude Desktop
Add to %APPDATA%\Claude\claude_desktop_config.json (Windows) or ~/Library/Application Support/Claude/claude_desktop_config.json (macOS):
{
"mcpServers": {
"kuka-krl": {
"command": "npx",
"args": ["-y", "kuka-krl-mcp"]
}
}
}4. Claude Code CLI
claude mcp add kuka-krl npx -y kuka-krl-mcpAvailable MCP Tools
Tool | Tier | Description |
| 0 | Looks up KSS system variables ( |
| 0 | Decodes KSS error numbers (e.g. 1376, 67, 2858) into root causes and SmartPAD step-by-step remediation procedures. |
| 0 | Provides language reference and code examples for PTP, LIN, CIRC, SPLINE, INTERRUPT, TRIGGER, FOLD, WAIT, CONTINUE. |
| 0 | Fast offline static linter: detects uninitialized $TOOL/$BASE, infinite loops, div-by-zero, unbalanced folds, and out-of-bound I/O. |
| 1 | Full project diagnostics across all |
| 1 | Discovers and lists all KRL modules in the active workspace. |
| 1 | Reads workspace KRL program text for contextual AI reasoning. |
| 2 (Pro) | Advanced industrial safety analysis: subprogram advance run hazards, payload verification, actuator-sensor races. |
| 2 (Pro) | Automatically fixes BOM/encoding, advance run hazards, fold desyncs, and trailing spaces. |
| 2 (Pro) | Generates Control Flow Graphs and Mermaid flowchart diagrams. |
| 2 (Pro) | Real-time operational telemetry from physical or virtual KUKA controller (TCP coordinates, axis angles, override, SmartPAD alarms). |
| 0 | Diagnostic tool checking local extension bridge connection and license state. |
MCP Prompts
krl-expert-review: Guides the AI model to perform a rigorous industrial safety review according to KSS standards.krl-diagnose-alarm: Guides the AI to analyze alarm codes from SmartPAD screenshots or logs and generate corrective action plans.
MCP Resources
krl://system-vars: Machine-readable JSON catalog of core KSS system variables.krl://syntax-guide: Complete KRL language keyword and syntax reference.
Unlocking Workspace Intelligence & Live Bridge
To connect kuka-krl-mcp with your project files and live robots:
Install the official KUKA KRL Extension for VS Code / Cursor.
Open your KUKA project folder.
The extension automatically starts a secure local loopback bridge on
127.0.0.1and authenticates your AI assistant.
License
MIT © LiskinLabs. Developed by Silvestr Liskin. Industrial robotics and automation solutions.
Available Tools
12 toolskrl_bridge_statusB
Checks connection status between this MCP server and the LiskinLabs KUKA KRL extension local bridge.
Inputs: no named parameters.
krl_control_flowA
Generates an execution Control Flow Graph (CFG) and Mermaid diagram for a KRL routine. Requires KUKA KRL Professional license.
Inputs: code (required).
krl_deep_safety_auditA
Performs deep industrial robot safety and motion verification (subprogram advance run hazards, payload verification, actuator-sensor races, turn unwind risks). Requires KUKA KRL Professional license.
Inputs: code (required).
krl_lint_snippetA
Fast offline static code analysis for KRL code snippets. Detects uninitialized $TOOL/$BASE, infinite loops without EXIT, literal division by zero, unbalanced ;FOLD/;ENDFOLD blocks, physical I/O bound overflows, and non-ASCII characters.
Inputs: code (required).
krl_live_telemetryA
Reads real-time operational telemetry from connected physical or virtual KUKA Robot Controller (robot model, KSS version, actual TCP position, joint angles, current line pointer, override $OV_PRO, mode T1/T2/AUT/EXT, active SmartPAD error alarms). Requires KUKA KRL Professional license and active controller connection.
Inputs: no named parameters.
krl_lookup_kss_messageA
Resolves KUKA KSS error and status message numbers (e.g. 1, 14, 20, 67, 68, 71, 72, 1376, 1422, 2858) into localized explanations, root causes, and industrial troubleshooting remediation steps.
Inputs: message_id (required), locale.
krl_lookup_system_varB
Looks up a KUKA System Software (KSS 8.3-8.7) system variable ($TOOL, $BASE, $OV_PRO, $VEL, $ACC, $POS_ACT, $IN, $OUT, $TIMER, $ADVANCE) with description, type, range, read/write permissions, syntax and examples in 6 languages.
Inputs: var_name (required), locale.
krl_read_fileA
Reads the text content of a KRL program or data file from the active IDE workspace. Requires LiskinLabs KUKA KRL extension.
Inputs: path (required).
krl_safe_auto_repairB
Automatically repairs common KRL hazards (BOM/encoding corruption, advance run desyncs, unbalanced inline form folds, trailing whitespace, uncommitted payload) and outputs verified repaired code. Requires KUKA KRL Professional license.
Inputs: code (required).
krl_syntax_referenceB
Provides authoritative syntax definitions and examples for KUKA KRL statements (PTP, LIN, CIRC, SPLINE, INTERRUPT, TRIGGER, FOLD, WAIT, CONTINUE) in 6 languages.
Inputs: keyword (required), locale.
krl_workspace_diagnosticsA
Collects project-wide continuous KRL diagnostics across all open .src, .dat, and .sub files from the active VS Code / Cursor / Antigravity workspace (SRC-DAT pair integrity, CIRC 2-point geometry, physical I/O ranges, FOLD balance, duplicate GLOBAL DEFs). Requires LiskinLabs KUKA KRL extension.
Inputs: no named parameters.
krl_workspace_filesB
Lists all KUKA KRL robot programs (.src), data files (.dat), and submit interpreters (.sub) in the active workspace. Requires LiskinLabs KUKA KRL extension.
Inputs: no named parameters.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 12 tools
Each tool has a discernible scope: lookups (system vars, KSS messages), syntax reference, snippet lint, workspace diagnostics, safety audit, repair, and telemetry are textually distinct. The only mild overlap is in the static-analysis family (krl_lint_snippet vs krl_workspace_diagnostics vs krl_deep_safety_audit), but descriptions clearly separate offline-snippet, project-wide, and motion/payload-safety concerns.
All 12 tools share the krl_ prefix and snake_case formatting with descriptive multi-word suffixes, giving a predictable namespace. There is a minor style split between verb-first names (krl_read_file, krl_lint_snippet, krl_lookup_*) and noun-phrase names (krl_syntax_reference, krl_control_flow, krl_live_telemetry), but readability and consistency remain high.
Twelve tools is a well-scoped set for a KRL development/assistant server, with each tool covering a distinct capability (connection, telemetry, references, analysis, repair, workspace access). No filler or redundant tools are apparent.
The surface covers the core read/analyze/repair lifecycle plus live controller telemetry and reference lookups, which is broad for the domain. Gaps remain: there is no write/save-file or create-program tool (krl_safe_auto_repair only returns repaired text), and no cross-workspace symbol search, but agents can work around these.
Maintenance
Related MCP Connectors
Connect Claude, Cursor, Codex, and other AI tools to your robotics mission data.
Read and write KukGit repositories, files, issues and pull requests from an AI assistant.
The Cortex MCP server provides read-only access to real-time engineering context from the Cortex developer portal, allowing AI coding assistants to answer natural language questions about your organization's catalog (microservices, libraries, domains, teams, infrastructure), scorecards (engineering standards and best practices), initiatives (goals and deadlines), and Engineering Intelligence metrics. It includes tools for querying documentation, tracking personal entities, and accessing AI-assisted insights across the entire Cortex ecosystem.
Bring your Knacco library into your AI tool: browse, search, read guides and download files.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceEnables AI assistants to interact with B\&R Automation Studio projects by building code, running ARsim simulators, and reading or writing OPC UA variables. It facilitates industrial automation development and real-time variable integration through natural language commands.17MIT
- FlicenseNot gradedqualityDmaintenanceEnables AI-powered PLC programming with natural language to ladder logic conversion, L5X and .ACD project generation, and semantic search through L5X files and documentation.43-
- FlicenseNot gradedqualityCmaintenanceEnables AI agents to connect to Allen-Bradley, Siemens, and CODESYS PLC projects for discovering, asking questions about, analyzing, and proposing code changes, with access to live tag values and version exports under permission-scoped controls.-
- AlicenseNot gradedqualityBmaintenanceEnables AI assistants to automate CODESYS V3 PLC development, including project editing, builds, library and alarm management, online variable read/write/force operations, and CODESYS documentation search via MCP.4MIT