An MCP server with OAuth2 authorization and remote agent capabilities, demonstrating OAuth provider integration and MCP tool registration on Cloudflare Workers
Single tool 'add' with basic schema and minimal documentation. Tool is clearly named (verb_noun pattern) and has a valid input schema with type definitions, but lacks depth in descriptions. The tool description is only 29 characters ('Adds two numbers together'), which falls below the recommended 50-200 character range for LLM optimization. Parameter descriptions exist but are minimal (13 chars for 'a', 12 chars for 'b'). No output schema is documented. No error handling guidance, recovery patterns, or context about when/why to use this tool. The implementation via Zod schema is correct, but the metadata layer is sparse for production use.
Adds two numbers together
Tool description is too brief (29 chars), below 50-char minimum for LLM clarity. Should explain: WHAT it does, WHEN to use it, and WHAT it returns. Current description 'Adds two numbers together' assumes context the LLM may not have.
Parameter descriptions are too short to be useful ('First number to add' = 19 chars, 'Second number to add' = 19 chars). Should include format constraints, allowed ranges, and use cases. E.g. 'First number to add (integer or decimal, range: -1e308 to 1e308)'.
No documented output schema. Tool returns text content via MCP response, but the description does not specify what format or structure the result will have. LLM cannot plan downstream steps without knowing output shape.
No error handling or recovery guidance. If tool receives non-numeric input or overflow, what happens? No error messages documented to guide LLM self-correction.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | D | 50 | 2026-07-28+ | v2 |
| 2026-03-09 | F | 23 | - | v1 |
No input validation constraints visible in schema or description. Zod validates at runtime, but LLM has no pre-call visibility into allowed types, ranges, or edge cases (e.g. Infinity, NaN, very large numbers).