caldera-mcp

Debug Ignition Perspective view issues including broken bindings, missing data, Bad_NotFound tag quality, silent component failures, and layout problems. Use when a view isn't displaying correctly, data isn't flowing through bindings, components behave unexpectedly, or the user mentions any Perspective debugging scenario. Also covers tracing transform chains, checking tag paths, diagnosing expression bindings, and using screenshots or console errors to identify problems. If something looks wrong in a Perspective view, this skill applies.

Explore and navigate Ignition gateway project structure, resources, and data. Use when browsing views, scripts, named queries, UDTs, tags, or gateway event scripts. Use when the user asks "what exists", "where is", "how is this structured", "show me", or wants to understand project organization. Activates for any Caldera MCP exploration task - listing resources, reading views or scripts, browsing tags, checking project overview, or searching for components. Even if the user just says "look at the project" or "what's on this gateway", this skill applies.

Plan new Ignition Perspective features by analyzing existing gateway patterns, component schemas, and design guidance. Use when the user has a feature request, user story, or wants to build something new - a new screen, dashboard, equipment detail view, alarm display, or navigation structure. Covers exploring existing conventions, checking component availability via search_components and get_component_schema, ISA-101/HMI design guidance, verifying data sources, and proposing view hierarchies grounded in what actually exists on the gateway.

Safety guardrails and write operation procedures for Ignition gateway modifications. Activates automatically before any write, delete, or modify operation on views, scripts, named queries, UDTs, tags, pages, or gateway event scripts. Covers the write checklist (read first, validate, surgical edits, verify after), QA vs production server rules, backup awareness, snapshot procedures, and read-only mode behavior. If any Caldera MCP write tool is about to be called, this skill provides the safety context.

Write and execute Jython 2.7 scripts on the Ignition gateway via execute_script and script sessions. Use when the user needs to probe databases, inspect tags, test expressions, run named queries, or perform any gateway-side investigation via scripting. Covers Jython 2.7 syntax rules (no f-strings, no walrus operator, no type hints), bridge execution context limitations, the _result return pattern, and script session management for iterative debugging. If the user needs to run code on the gateway or you need to test a hypothesis by executing a script, this skill applies.