svelte-component-graph-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| get_graphA | Return the full Svelte component dependency graph for a project: every component/route node (workspace-relative path id + display label + type + unused flag) and every import link. |
| get_componentA | Look up a single component or route by its workspace-relative path (e.g. "src/lib/Button.svelte"). Returns its parents (components that import it), children (components it imports), whether it is unused/a route, and its Svelte 5 public API surface: props (name, optional, bindable, rest) and slots. |
| get_unusedA | Return every component that is imported somewhere in the project but rendered nowhere (never referenced in any importer's template). Computed globally, so the result does not depend on file order. |
| get_routesA | Return every route node (+page/+layout/+error) with its display label and the direct child components it pulls in. |
| scanA | Invalidate all caches for the project and re-parse every file from scratch. Returns a summary with the node and link counts. Use when you want to guarantee fresh results. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 5 tools
Each tool targets a distinct aspect of the component graph: global view (get_graph), per-node detail (get_component), unused analysis (get_unused), route view (get_routes), and cache refresh (scan). There is no functional overlap.
Four tools consistently follow the 'get_' prefix pattern, while 'scan' is a bare verb, which is a minor deviation but still understandable.
With five tools, the server strikes a good balance—neither too sparse nor overwhelming for its focus on graph analysis.
The tools cover all necessary operations for a read-only component graph server: fetching the full graph, drilling into a component, finding unused components, routing information, and cache invalidation. No obvious missing capability.