Loading ToolBench
MCP server providing tools for product, user, commodity, role, currency, vendor, and Jira issue management
This server has 21 tools with consistent naming patterns and basic parameter schemas, but several quality gaps prevent a higher score. All tools follow verb_noun naming (get-, search-, create-) which is good. However, most parameter descriptions are minimal (10-30 chars), and output schemas are completely undocumented. Error handling is not visible in the source. The two write operations (create-jira-issue, create-jira-subtask) lack confirmation/dry-run patterns. Tool descriptions are adequate but generic (50-80 chars), missing context about when to use each tool vs. similar ones. The 21-tool surface area with repetitive patterns (get-X, search-X, get-X-by-id across 5 resource types) suggests composition gaps, users likely need multi-step chains that aren't pre-composed.
Create a new Jira issue with the provided data
Create a sub-task for a given Jira issue
Get a list of all product categories
Get a list of commodities with optional pagination
Get detailed information about a specific commodity by its ID
Get a list of currencies with optional pagination
Get detailed information about a specific currency by its ID
Output schemas completely undocumented, tool descriptions do not specify what fields are returned, their types, or what IDs/references are available for chaining
Parameter descriptions are minimal (10-30 chars) and lack context. E.g., 'limit' desc is 56 chars but does not explain default behavior or explain when to use pagination. Most descriptions fail to answer WHEN to use this tool vs. similar ones.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | B | 74 | 2026-07-28+ | v2 |
| 2026-03-09 | D | 53 | - | v1 |
Get detailed information about a specific Jira issue by its ID
Get detailed information about a specific product by its ID
Get a list of products with optional pagination
Get all products in a specific category
Get detailed information about a specific role by its ID
Get a list of roles with optional pagination
Get detailed information about a specific user by its ID
Get a list of users with optional pagination
Search for commodities by keyword
Search for currencies by keyword
Run a custom Jira JQL query to search for issues
Search for products by keyword
Search for roles by keyword
Search for users by keyword
Destructive operations (create-jira-issue, create-jira-subtask) lack dry-run or confirmation patterns. Agents cannot preview changes or be gated by a confirmation step before creating Jira issues.
Repetitive tool proliferation: 15 of 21 tools follow three patterns (get-X, get-X-by-id, search-X) repeated across 5 resource types. No single tool composes common workflows (e.g., search product by category + get details). Agents must chain multiple calls for simple user intents.
No error handling or recovery guidance visible. Source shows no try-catch patterns, no validation messages, no suggestions for alternatives when a resource is not found.
Parameter descriptions do not include constraints (min/max, regex, enums) textually, they rely on JSON Schema. LLMs cannot reliably read JSON Schema from descriptions alone. E.g., 'limit' description should state '(1-100, default 20)' in the description text itself.
Tool descriptions do not clarify distinguishing factors. E.g., get-products vs. get-products-by-category both retrieve products, the description should explain when each is preferred. search-products uses keyword search; get-products-by-category uses category filtering, this distinction is not made explicit.