A two-server MCP system for PCB assembly process orchestration and retrieval-augmented generation (RAG) over process documentation. The orchestrator server manages PCB assembly process sequences with validation and execution, while the RAG server provides semantic search over PCB assembly process documentation.
Server has 9 tools with basic descriptions and schemas, but significant quality gaps. Tool names lack clear verb-noun patterns (e.g., 'set_sub_param_at_position' is vague about what 'sub_param' means). Descriptions are present but generic, most lack context on WHEN to use each tool or what happens on failure. Parameter descriptions are minimal; 'sub_param' in tool 1 lists valid values inline rather than as enums. Output schemas are undocumented, callers don't know what structure to expect. Error handling is absent; no guidance on recovery or retryability. Tool 4 (execute_process) is irreversible but has no confirmation pattern. Tool 9 (get_query_rag) is incomplete in source. Overall, definitions are functional but fall short of production quality.
Execute the current process if it's valid. Returns: Execution status message indicating success or failure
Get valid sub-parameters for a specific block type. Args: block_type: Block type to query Returns: List of valid sub-parameters for the specified block type
Get the current state of the orchestrator process. Returns: Current process with all blocks and parameters, plus validation status
Check if the current process is valid. Returns: Validation status and which process it matches (if valid)
Get all possible block types and their valid sub-parameters. Returns: Formatted string listing all block types and their sub-parameters
Search the RAG system for general information about PCB assembly processes. Args: query: The search query about PCB processes, materials, or techniques Returns: Relevant information from the documentation
Parameter 'sub_param' in set_sub_param_at_position lacks enum constraint; valid values listed inline in description instead of as formal schema enum. LLMs will hallucinate invalid values.
Output schemas undocumented for all tools. Callers don't know what fields to expect (e.g., does get_current_process return JSON or plain text? What fields?). Breaks downstream tool chaining.
execute_process is irreversible (WRITE risk) but has no confirmation step, dry-run option, or error recovery guidance. Agents can accidentally trigger production assembly without safeguards.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 47 | 2026-07-28+ | v2 |
Get all valid process processs that the orchestrator accepts. Returns: Formatted string listing all 9 valid processs with their steps
Set a sub-parameter at a given position. Args: pos: Position (0-4) in the sequence sub_param: Sub-parameter value. Valid values depend on the block type: - Solder Paste Application: 'lead-free', 'leaded', 'low-temp' - Component Placement: 'high-speed', 'high-precision', 'flexible' - Soldering: '235C', '245C', '260C' - Optical Inspection: '2D', '3D', 'Automated' - Testing: 'in-circuit', 'functional', 'boundary-scan' Returns: Status message with current sequence
Set the entire process sequence with sub-parameters. Args: sub_params_sequence: List of 5 sub-parameters for each block in order. Returns: Status message with current sequence
Tool names lack clear verb-noun patterns. 'set_sub_param_at_position' is ambiguous, does it set one param or multiple? 'get_possible_blocks_sub_params' is verbose. Rename to 'set_sub_parameter', 'list_block_parameters'.
No error handling or recovery guidance. Tools return plain strings; no structured error responses. If set_sub_param_at_position fails, LLM gets a string error with no actionable next step.