Next.js AI client application with MCP server integration, chat interface, and artifact management
Two tools with partial schemas and minimal descriptions. manage_artifact has a schema but lacks parameter descriptions and output documentation. search_knowledge_base has only a query parameter with minimal context. Neither tool description explains when to use it, what it returns, or error handling. No tool annotations (readOnlyHint/destructiveHint). Descriptions are 80-90 chars (below 194-char baseline). Parameter descriptions are absent or trivial. No pagination, error guidance, or output schema documentation visible.
Allows the AI to generate interactive Markdown, Spreadsheets, HTML UI, and Mermaid diagrams in a side panel.
Searches the knowledge base using hybrid search to retrieve relevant documents and content.
Parameter descriptions missing. manage_artifact has 4 parameters (type, title, content, sheets) with no descriptions explaining what each controls, valid values, or constraints. LLMs cannot infer parameter meaning from names alone.
Output schemas not documented. Neither tool specifies what fields are returned, their types, or structure. LLMs cannot plan downstream calls or extract data without knowing response shape.
Tool descriptions too brief and lack context. manage_artifact (80 chars) and search_knowledge_base (90 chars) do not explain WHEN to use each tool, what it returns, or prerequisites. Baseline is 194 chars.
No error handling guidance. Neither tool describes what errors can occur, how to recover, or what the LLM should do next. Agents have no recovery path on failure.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 45 | 2025-06-18+ | v2 |
Missing tool annotations. No readOnlyHint or destructiveHint declared. search_knowledge_base should be marked readOnly; manage_artifact should be marked destructive (modifies state).