Loading ToolBench
K3B MCP has basic tool definitions with schemas present, but suffers from significant gaps in parameter descriptions, output documentation, and error guidance. All 4 tools have input schemas with proper types and required fields, but descriptions are minimal (10-80 chars, well below the 194-char baseline). Parameter descriptions are sparse or missing context. No output schemas documented. Error handling returns raw exit codes without recovery guidance. The run_k3bcli tool is a dangerous escape hatch that bypasses safety constraints.
Create and burn a data disc from a folder. Uses k3bcli data project mode.
Show k3bcli help output to discover supported flags and modes.
List available optical drives using k3bcli.
Run k3bcli with arbitrary arguments. Use this when you need flags not covered by the other tools.
run_k3bcli tool is a dangerous escape hatch that bypasses safety constraints. Allows arbitrary k3bcli arguments without validation, enabling destructive operations (erase, format) without confirmation or audit trail.
No output schemas documented. LLMs cannot predict what fields to expect from tool responses, forcing them to parse unstructured text and preventing downstream tool chaining.
Parameter descriptions lack actionable constraints. 'Optional burn speed (e.g. 4, 8, 16)' uses examples instead of formal constraints. No min/max bounds, no enum for device paths, no validation rules stated.
Error responses provide no recovery guidance. When k3bcli fails, the tool returns raw stderr without telling the LLM what to do next (retry? check device? try different flags?).
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 46 | 2026-07-28+ | v2 |
Tool descriptions are too brief (10-80 chars vs 194-char baseline). 'Show k3bcli help output' and 'List available optical drives' lack context on WHEN to call them or what they return.