iDesktopX Desktop Agent Assistant
Paste the following prompt into your AI chat to install this skill:
Please install @user_f47c39e1/idesktopx-skills according to https://skillhub.cn/install/skillhub.md.
About this skill
The problem it addresses
When agents operate SuperMap desktop products directly, the risk is not simply whether a control can be clicked, but whether the command entry point is correct, the product version is compatible, REST is already enabled, and success=true has been confused with business completion. idesktopx-skills is built for this desktop automation path. It binds the target product root, the DesktopCLI wrapper, online HTTP, and CLI fallbacks into the same session. It can target same-origin products such as iDesktopX, TransformX, ImageX, iMaritimeEditor, and ActiveMap, while the actual root directory must come from the user.
How it works
The skill enforces capability routing: prefer the semantically closest domain skill, such as tree/*, form/*, map/*, or dataset/*, and fall back to component/*, dialog/*, or the Python sub-skill only when a dedicated domain capability is missing.
Before the first business operation in a session, it requires several gates:
- Read optional user-level config, or ask for the target product root
- Bind the
idesktopx/idesktopx.cmdabsolute path under that root - Call
desktop/version_getand validate the12.1.version prefix - Reuse REST through
rest status; runrest enableonly when disabled - Prefer HTTP endpoints such as
GET /api/v1/healthandPOST /api/v1/{domain}/{action}
During execution, it emphasizes minimal probing and read-back validation: after form/activate, verify form/active; after table/setvalue, verify table/getvalue; treat GPA asynchronous tasks as complete only when gpa/status reaches a terminal state. When BLOCKED_DIALOG appears, it stops business commands, reads the blocking window and component structure, and handles user-decision dialogs carefully.
Boundaries
It is best suited for automation against a running desktop product instance, not for modifying or deleting source data without authorization. Configuration is a cache, not a prerequisite. If the version is mismatched, the root is ambiguous, or the REST endpoint cannot be read, the skill should stop and confirm instead of switching to another same-named command on PATH.
Use Cases
- After the desktop product starts, check REST and version gates, then call map or layer domain commands and read back state.
- When reading dataset records, use `dataset/query_records` paging instead of exporting the full table as CSV by default.
- When a modal dialog blocks automation, read the window and component structure, act only with authorization, then re-read state.
- When external Python accesses desktop runtime objects, load the Python sub-skill, detect the gateway, and read datasources or maps.
Best For
- GIS engineers maintaining SuperMap desktop automation who need controlled CLI/REST calls for maps, layers, and datasets.
- Product engineers building desktop agents who need stable gates for version matching, REST status, and blocking dialogs.
- Business engineers using GPA or spatial analysis who need to rely on `gpa/status` terminal states rather than output windows.
- Data engineers writing external Python scripts who need to connect to the desktop runtime and read workspace, datasource, dataset, or map objects.
Related Skills
Automatically indexes Gradle-cached AAR/JAR dependency classes and returns library coordinates, versions, and public APIs by fully qualified name, using only the Python standard library.
Codifies AMT and YourMT3 training conventions, script patterns, hyperparameters, precision, checkpoints, and NaN safeguards.
Retrieve relevant chunks from a customer-managed PKM dataset by dataset_id and return concise, source-annotated answers.
Convert PRDs, user stories, or functional specs into prioritized test-point checklists covering functional, business-rule, boundary, exception, and non-functional dimensions.