This server exposes only 1 tool (get-custom-elements) with a READ_ONLY risk profile. The tool name follows verb-noun convention (get_*), which is positive. However, critical definition issues prevent a higher score: (1) The input schema is not visible in the provided source code, only the tool name and description are shown. (2) No parameter descriptions are evident. (3) The tool description ('Get all custom elements and web components available in the workspace') is generic and under 50 chars, lacking guidance on when to use it, what the output structure is, or what fields the LLM should expect. (4) No documented output schema is visible, LLMs cannot reason about what fields will be returned or how to chain this tool to others. (5) No pagination, limit, or offset parameters visible despite the potential for large result sets. (6) The server is a VSCode extension exposing an HTTP MCP interface; the actual tool implementation is not shown in the provided code excerpt, making full verification impossible.

• Document the full input schema for get-custom-elements, including any optional parameters such as filter, limit, offset, or manifest_uri. Example: { 'type': 'object', 'properties': { 'filter': { 'type': 'string', 'description': 'Optional filter for element names (e.g., prefix match)' }, 'limit': { 'type': 'integer', 'minimum': 1, 'maximum': 100, 'default': 20, 'description': 'Max results to return' }, 'offset': { 'type': 'integer', 'minimum': 0, 'default': 0 } } }.
• Expand the tool description to 100 - 200 chars. Example: 'Retrieve all custom elements and web components detected in the workspace custom-elements.json manifests. Returns element names, properties, attributes, events, and slots. Use this to discover available components before generating code or asking the user which component to work with.'
• Document the output schema explicitly. What fields does each returned element have? Example output structure: { 'elements': [ { 'name': 'my-button', 'tagName': 'my-button', 'properties': [...], 'attributes': [...], 'events': [...], 'slots': [...] } ], 'total': 42, 'returned': 20, 'nextOffset': 20 }.
• Add pagination support: return a total count and next offset/cursor so agents can fetch large element lists without blowing context. Cap results at 20 - 50 per call.
• Add error handling guidance. Extend the tool to return clear error messages: 'No custom-elements.json found in workspace. Create one or ensure web component manifests are in scope.' or 'Parsing error in manifest at <path>: <details>. Check JSON syntax.'

Inferred effective spec: <=2025-11-25.

• Relies on Logging (deprecated) - log to stderr or use OpenTelemetry