Conduit's test fixture MCP server defines 3 tools with minimal quality. All tools have descriptions (10-60 chars, below the 194-char baseline), but lack parameter descriptions entirely. Input schemas are present with proper JSON Schema structure (type, properties, required), but parameters lack descriptions, a critical gap per the rubric. No output schemas are documented. Error handling is absent (the 'fail' tool throws but provides no recovery guidance). Tool names are verb-based (echo, add, fail) but generic; 'fail' is particularly vague. No tool annotations (readOnlyHint, destructiveHint, idempotentHint) are present. This is a minimal test fixture, not a production-grade server.

• Add descriptions to all parameters. E.g., 'text' → 'The input string to echo back verbatim (1-10000 chars)'; 'a' and 'b' → 'A number to add (integer or float, range -1e9 to 1e9)'.
• Expand tool descriptions to 100-200 chars. E.g., 'echo' → 'Returns the input text verbatim. Useful for testing tool invocation and verifying parameter passing. Returns the exact string provided.'
• Document output schemas. E.g., echo returns {type: 'object', properties: {result: {type: 'string'}}, required: ['result']}; add returns {type: 'object', properties: {sum: {type: 'number'}}, required: ['sum']}.
• Add error handling to 'fail'. Instead of throwing, return {error: 'Intentional failure', message: '<user message>', recoveryHint: 'This tool is designed to test error paths. Retry with a different message or skip this step.'}.
• Add tool annotations. Mark 'echo' and 'add' with readOnlyHint: true; mark 'fail' with destructiveHint: false (it does not modify state, only tests error paths).