Loading ToolBench
A Model Context Protocol server for querying AWS Bedrock Knowledge Bases
Single tool with adequate schema and clear description, but lacks depth in parameter documentation and output schema specification. Tool name follows verb_noun pattern (query_knowledge_base). Description is 127 chars, within acceptable range. Input schema is present with proper types. However, output schema is not documented, LLMs cannot predict response structure. Parameter descriptions are minimal (e.g., 'temperature' lacks guidance on how it affects output). No error handling guidance or recovery paths documented. No tool annotations (readOnlyHint, etc.). Composition is simple but functional.
Query the AWS Bedrock Knowledge Base to retrieve and generate relevant information
Output schema not documented. LLMs cannot predict response structure or plan downstream operations. Tool returns text in a content array, but this is not declared in tool definition.
Parameter 'temperature' lacks actionable guidance. Description states range (0.0-1.0) but does not explain effect: 'Controls response randomness. Lower (0.0) = deterministic; higher (1.0) = creative. Default 0.7 balances both.'
No error handling guidance. Tool catches errors and returns isError=true with message, but does not guide LLM on recovery. E.g., 'If authentication fails, verify AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY are set.'
No tool annotations. Missing readOnlyHint (this is a read-only query tool) and idempotentHint (repeated calls with same query return same result). These hints help agents reason about retry safety and side effects.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 52 | 2026-07-28+ | v2 |
Parameter 'context' description is generic. Should explain when to use it: 'Optional background to narrow results. E.g., if querying about EC2, provide context="We use t3.medium instances in us-east-1" to get region-specific guidance.'