Loading ToolBench
Anonymous read-only scientific-figure workflow knowledge server. Provides templates, rules, schemas, and version metadata for scientific figure creation, refinement, and evaluation workflows.
Single tool with well-structured schema and clear enum constraints. Tool name follows verb_noun pattern (get_workflow_bundle). Input schema is complete with proper types, enums, and descriptions for all 5 required parameters. Description is concise (106 chars, within 10-1024 baseline). However, output schema is not documented in the tool definition itself, only visible in the knowledge package exports. No error handling guidance or recovery patterns documented. Tool is read-only (good security posture) but lacks explicit idempotent/readonly annotations in the MCP tool definition.
Returns public templates, rules, schemas, and version metadata for one fixed workflow selection.
Output schema not documented in tool definition. LLMs cannot infer what fields the response contains or plan downstream operations.
No error handling guidance. Tool definition lacks recovery hints for invalid inputs (e.g., unsupported locale or visualCategory).
Tool annotations missing. readOnlyHint should be set to true in tool definition to signal to LLMs that this tool is safe to call without side effects.
knowledgeMajor parameter is over-constrained with const:1. This prevents future versioning and should use enum:[1,2,...] or be removed if always 1.
Inferred effective spec: 2026-07-28+.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | C | 66 | 2026-07-28+ | v2 |