Loading ToolBench
A professional MCP server for managing workspaces and shared instructions with a 5-layer architecture
Server has 6 tools with complete input schemas and descriptions present. Naming follows verb_noun convention (create_, list_, get_, update_). However, descriptions are generic and lack LLM-optimized guidance on WHEN to use each tool or what distinguishes similar tools. Parameter descriptions exist but are minimal (avg ~30 chars vs baseline 72 chars). No output schemas documented. Error handling is absent, no recovery guidance or actionable error messages. Tool composition is reasonable but lacks idempotency guarantees and chaining metadata.
Create a new shared instruction template
Create a new workspace with optional description and template
Get detailed information about a specific workspace
List all available shared instruction templates
List all available workspaces
Update global instructions that apply to all workspaces
No output schemas documented. LLMs cannot infer what fields to expect from tool responses, forcing them to guess at downstream chaining and risking context loss.
Descriptions are generic and lack LLM-optimized guidance. 'Create a new workspace with optional description and template' does not explain WHEN to use this vs alternatives, what happens on success, or what IDs are returned for chaining.
No error handling or recovery guidance. Tool implementations do not document what errors can occur, whether they are retryable, or what the LLM should do next.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 56 | <=2025-11-25 | v2 |
Parameter descriptions are minimal (avg ~30 chars vs baseline 72 chars). E.g., 'The name of the workspace' lacks context on format, length, allowed characters, or uniqueness constraints.
No idempotency guarantees documented. Agents retry on ambiguous failures, create_workspace and create_shared_instruction should clarify whether repeated calls with identical inputs are safe or risk duplicates.