Loading ToolBench
A monorepo containing multiple MCP servers including an AI phone assistant and a biography provider server
Both tools have empty input schemas (no properties defined) and minimal descriptions. Tool names lack action verbs (biography_provider, load_biography are noun-heavy). No parameter descriptions exist. Schemas are present but trivial. No error handling guidance, no output schema documentation, no security considerations. This server falls well below production baseline.
Provides a biography of a Vlad Krasovski
Vlad Krasovsky initial biography and summary if his personal and professions background
Empty input schemas: both tools declare input as {"type":"object","properties":{}} with no parameters.
Tool names lack action verbs. 'biography_provider' and 'load_biography' are noun-heavy and do not clearly signal what action occurs. Should be 'get_biography' or 'fetch_biography' to match verb_noun convention.
Descriptions are under 50 characters and lack context. 'Provides a biography of a Vlad Krasovski' (typo: Krasovski vs Krasovsky) and 'Vlad Krasovsky initial biography...' do not explain WHEN to use, WHAT it returns, or any prerequisites.
No output schema documented. Callers cannot know what fields to expect from the response. biography_provider returns {content: [{type: 'text', text: biography}]} but this is not declared in the tool definition.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 29 | 2026-07-28+ | v2 |
Duplicate/overlapping tools. Both biography_provider and load_biography appear to do the same thing (fetch biography). LLMs will waste reasoning cycles deciding between them. Consolidate into a single canonical tool.