Loading ToolBench
Two tools with minimal schema and weak descriptions. initialize_connection has empty input schema ({}), execute_query has a single string parameter with no type validation or constraints. Descriptions are present but lack actionable guidance for LLM selection. No parameter descriptions, no output schemas documented, no error handling guidance. Tool names are reasonable (verb_noun pattern) but descriptions do not explain WHEN to use each tool or what to expect. No pagination, no structured output specification, no recovery hints.
Executes a query on the database. Make sure that you lower case the text and the column compared and use the LIKE operator. This is very important.
Initializes connection to Chakra and provides the user with the database metadata.
initialize_connection has empty input schema ({}). Tool provides no way for LLM to understand what inputs it accepts or requires.
execute_query accepts 'query' parameter with no type validation, format constraints, or description of expected SQL dialect. LLM cannot infer valid syntax or guard against injection.
No output schemas documented for either tool. LLMs cannot plan downstream operations or extract structured data. initialize_connection returns metadata as string; execute_query returns DataFrame.to_string(). Both are unstructured.
Error handling is minimal. execute_query returns a plain string error if not connected, but does not guide LLM on recovery (e.g., 'Call initialize_connection first'). No categorization of retryable vs fatal errors.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 42 | 2026-07-28+ | v2 |
execute_query description includes implementation detail ('lower case the text and the column compared and use the LIKE operator') rather than explaining WHAT the tool does and WHEN to use it. Mixes instruction with purpose.