apt-marketplace-test

Manage baselines and change history in a Sparx EA repository via ea_repository and ea_analyze — when to baseline (whole model vs. one package), how to read a compare_baseline diff, the destructive behavior of apply_baseline and how to protect against it, when to proactively offer a baseline before a significant change (package delete, large bulk edit, ruleset fix, MDG rollout, XMI import), how to review accumulated baselines by payload size and guide cleanup, and using get_updates_in_range / get_user_activity to answer "what changed, when, and by whom." Use before a risky bulk edit or MDG/ruleset rollout, when asked to snapshot, diff, or roll back model state, when a baseline should be offered before a destructive operation, when reviewing or pruning old baselines, or when investigating recent changes or an editor's activity in a Sparx EA model.

Use the Sparx EA COM API from Python to read and modify EA models — connect to a running instance, run SQL queries, manage MDG technologies, create/update elements, and control EA lifecycle. Use when automating EA work from a Python script.

Generate a structured diagnostic report when a customer encounters issues with AI Power Tools for Sparx EA, then guide submission to [email protected]. Invoke when a user asks for help with an issue, when an unexpected error occurs that cannot be resolved in the session, when a workaround was applied but the underlying bug should be reported, or before ending a session where multiple issues were encountered.

Recommend which Enterprise Architect diagram type and viewpoint expresses what a user wants to understand — from their intent, the measured content of a scope, and the languages that content actually uses. Handles the hard case where a scope spans several modeling languages and no conformant view exists. Use when someone describes what they want to see rather than naming a diagram type, or before generating a diagram over content whose modeling languages you have not checked.

Compose Enterprise Architect diagrams that read as deliberate — choose a layout grammar, compute geometry, place and style elements and connectors, then verify and correct the result. Language-neutral; works with any modeling language, reading each notation's conventions from a data binding (several notations ship). Use when building a diagram from model content rather than hand-placing elements, or when a generated diagram looks untidy and you need to know why.

Fallback route for a Sparx EA task no other skill covers — identify the running EA version and repository type, open Sparx's user guide for that exact version, find the written procedure, and carry it out in EA's UI. Use only after ea-start-here has found no skill for the task, or when the task has no MCP operation and no COM route. Never use it in place of a skill that already covers the job.

Get content into and out of a Sparx EA repository — XMI package export/import, diagram export to PNG/SVG/Visio, and RTF report generation via ea_repository and ea_diagram. Use when a package or diagram needs to leave EA (another repository, a document, a stakeholder without an EA license) or an XMI file needs to come back in.

Work out what modeling-language situation a Sparx EA repository is actually in — ad hoc stereotypes vs. a client-built MDG, and whether a technology is installed, embedded, or actually loaded — before picking ea-mdg-author or ea-mdg-model-build. Use at the start of any MDG task, when a stereotype behaves inconsistently, or whenever assess_mdg_situation's recommendation needs a second opinion.

Author a Sparx EA MDG Technology XML file — define stereotypes, tagged values, toolbox pages, and custom diagram types. Use when creating or modifying any MDG technology for Sparx EA.

Deploy and test a Sparx EA MDG Technology — embed it into a .qea model file or install application-wide, then verify it works correctly. Use after authoring or modifying an MDG XML file.

Build a Sparx EA MDG Technology from a model held inside the repository — requirements, design metamodel, profile packages, toolbox and diagram profiles, then export and generate the MDG XML. Use when the MDG has a source model in EA, or when a hand-authored MDG needs to be brought under model control.

Find and safely fix model decay in a Sparx EA repository -- elements with no connectors, composite-diagram links that fell out of sync, and stereotypes or tagged values applied inconsistently. Use when a model has been growing for a while and needs a cleanup pass, before deleting or moving anything that might still be load-bearing, or when auditing a repository for drift from its MDG.

How to build and maintain Sparx EA models through the ea-mcp-server MCP tools — correct build order, known defects and workarounds, verification patterns, and governance considerations. Use this skill whenever constructing or modifying an EA repository via MCP tool calls.

How to generate click-through "navigation diagrams" over hierarchical data in a Sparx EA repository — the composite-element mechanism, the nested-container layout, the exact write sequence, and verification. Use this skill whenever a repository contains or has just been loaded with hierarchical elements (capability models, taxonomies, org structures, product breakdowns, reference models) and the user wants to browse them by clicking down through levels rather than reading a tree diagram.

Put a Sparx EA repository into Power BI as a generated semantic model - a complete .pbip project with TMDL, declared relationships, one measure per inactive `Con_` relationship and EA's own tables hidden - reading the business layer that ea-reporting-database builds, whether as SQL views in the EA database (DirectQuery or Import), as views in a SQL Server reporting database (Import), or as Parquet files (Import). Use when someone wants EA content in Power BI, asks for a dataset, a .pbip, TMDL or a semantic model, asks which of the three paths to take, asks how to refresh, or asks why Power BI's own relationship detection does not give them what they expect.

Build the reporting layer for a Sparx EA repository - business tables named and shaped by the technology that governs the model (one table per stereotype, one per allowed connector combination, a Table Directory) over EA's own tables - as SQL views inside the EA database, as a SQL Server reporting database, or as Parquet files, and prove it matches the repository. Use when someone wants to query, roll up, or report across a whole model rather than element by element, wants EA content in a database or files that an agent, a SQL client or Power BI can read, wants to choose what is reportable, or asks to keep the data fresh on a schedule. For Power BI specifically - a dataset, a .pbip, TMDL or a semantic model - build the layer here, then continue with ea-power-bi.

Build a YAML validation ruleset for a modeling language (ArchiMate, BPMN, UML, SysML, custom MDG, etc.), consumed by validate_model. Covers the full authoring workflow: research → rule-category planning → YAML authoring → testing. Use this skill at the start of every new language ruleset, and keep it open throughout.

Turn a Sparx EA model into an answer a non-modeller can act on -- portfolio roll-ups by tagged value, plain-language element summaries, and business-facing display terms. Use when someone asks "how many X are Y", "what's mission-critical", "what's in scope for [a regulation]", or wants a model fact stated in business language instead of stereotype/tag names, and especially before quoting any count drawn from a tagged value.

Start here for any Sparx Enterprise Architect task. Runs a short session preflight, then routes to the right skill for the job — modeling, MDG authoring or deployment, validation, diagrams, or COM automation. Use this first whenever a session involves an EA model, before reaching for a more specific EA skill.

Author and run YAML sidecar conformance rules against a Sparx EA model via the ea_validate meta-tool. Use when you need to check that an MDG-modeled repository meets its required tagged values, connector cardinality, and endpoint stereotype constraints.

Build, read and change Prolaborate dashboards through the web interface — choose the right widget for a question, write the query each widget type requires, verify it before saving, and hand back a layout plan. Use when the task names a Prolaborate dashboard, widget, chart, card or report tile, or asks what a dashboard shows. Never use it for Enterprise Architect model authoring, which belongs to the EA skills.

Trace dependencies and analyze the impact of change in Prolaborate using Analyzer views — seed a graph from an element, expand it one hop at a time, keep it readable, and report what is actually affected. Use when the task asks what depends on something, what breaks if an element changes, how two elements are connected, or asks to build or read an Analyzer view. Never promise a complete blast radius — the product has no depth traversal, and §2 says what that means.

Read, filter, share and export Prolaborate relationship matrices through the web interface — find the right profile, join the response into a correct grid, and report what the cells mean. Use when the task names a relationship matrix, asks which elements are linked to which, or asks to export or share one. Never use it to create a matrix profile or to write matrix cells; both are Enterprise Architect work.

Start here for any Sparx Prolaborate task driven through the web interface. Runs a short session preflight — which tenant, which signed-in identity, which version, which repositories — then routes to the right Prolaborate skill for dashboards, relationship matrices or impact analysis. Use this first whenever a session involves Prolaborate, before reaching for a more specific prol- skill.