A sample MCP server demonstrating math operations, weather, and text processing tools
Server defines 4 tools with basic schemas and descriptions. All tools follow verb_noun naming convention (add, multiply, get_weather, count_words). Descriptions are present but minimal (8-42 chars), falling short of the 10-1024 char baseline. Parameter descriptions exist but lack detail on formats, ranges, or constraints. No output schemas documented. Error handling absent. Return type for tools is evident from code but not explicitly documented for LLM consumption. Tools show low token optimization, no pagination, no batching, no composition chains evident.
Add two numbers
Count the number of words in text
Get weather information for a location
Multiply two numbers
Descriptions are too brief (8-42 characters, baseline 10-1024). LLMs cannot distinguish when to select add vs multiply, or why get_weather exists vs other tools.
No output schemas documented. LLMs do not know what fields to expect from get_weather (string response) or count_words (integer return). This blocks downstream tool chaining and forces the LLM to guess the structure.
Parameter descriptions lack formatting constraints. The 'location' parameter in get_weather should specify format (city name, coordinates, country code?). The 'text' parameter in count_words should state expected length or content type.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | D | 50 | 2026-07-28+ | v2 |
| 2026-03-09 | D | 53 | - | v1 |
No error handling guidance. If get_weather receives an invalid location or count_words receives null input, there is no recovery instruction for the LLM. No mention of retryability or partial failure modes.
Mock weather data in get_weather is hardcoded. Description does not clarify that this is mock/demo data, not a real weather service. LLM may be confused about reliability or real-world utility.