Loading ToolBench
MCP server for Repetier-Host / Repetier-Server 3D printer monitoring, control and intelligent diagnostics
Server has 11 tools with basic descriptions and input schemas, but significant gaps in parameter documentation and output schema clarity. Most tool descriptions are adequate (50-150 chars) but lack actionable context for LLM selection. Parameter descriptions exist but are minimal. No visible output schema documentation. Error handling is present but generic. Tool naming follows verb_noun convention well (get_*, list_*, send_*, etc.), which is a strength. However, parameter validation rules and constraints are not documented in descriptions, and several tools lack guidance on when to use them vs. similar tools.
Cancel the currently printing job
Analyze printer errors and anomalies with intelligent diagnostics using Sidewinder X1 knowledge base
Query EEPROM settings from the printer
Get real-time printer status including temperatures, position, and job progress
Get historical temperature readings for hotend and bed
List all queued print jobs on the Repetier-Server
Pause the currently printing job
No output schemas documented for any tool. LLMs cannot plan downstream calls or extract required fields without knowing response structure.
Tools with no input parameters (pause_job, resume_job, cancel_job) lack descriptions explaining how they determine which job to target. Are they stateful? Do they operate on the currently printing job?
Parameter descriptions are minimal (e.g., 'The ID of the job to start'). Missing context on format, constraints, and when to use lookup tools first.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 50 | 2026-07-28+ | v2 |
Resume a paused print job
Send a raw G-code command to the printer
Start a queued print job by job ID
Upload a G-code file to Repetier-Server for printing
No error recovery guidance. Tools return generic error messages without suggesting next steps (e.g., 'Cannot connect to Repetier-Server' should suggest checking SERVER_URL or API key).
Destructive operations (cancel_job, send_gcode with dangerous commands) lack confirmation or dry-run support. Agents could accidentally cancel prints without safeguards.