Data Profile for This Category
Data sources for this category include ERP material master data, melting furnace thermometers, spectrometers, MES heat records, mold management systems, customer drawings, and technical specifications. Update frequencies cover per heat, per batch, and daily. Daily documents include foundry process cards, batching lists, heat records, mold histories, first article inspection records, material test reports, and nonconforming product review forms. Core fields include heat number, pouring temperature, mold model, material grade, raw material weight, and inspection values. Supported units include degrees Celsius, kilograms, millimeters, and others.
Constraints for Workflow Orchestration
The data for this category is scattered and updated frequently. Workflows must support pulling and integrating multi-source data. Configure routing logic for tool calls and MCP calls to match access requirements for different data sources. Multiple document types require workflow document upload nodes to support a wide range of formats. Core fields use specific units. Form input nodes must support variables global variable references to automatically populate standard parameters and reduce manual input errors. High-frequency heat-level updates require workflow nodes to pull the latest heat data in real time using trigger conditions, ensuring quotations and documents stay up to date.
Configuration Guidelines
This configuration is built for FastGPT v4.17.0.
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
| Form input node default value | Reference variables global variables or form field values | Adapt to dynamic parameter requirements of foundry categories, automatically populate fixed fields such as heat number and material grade |
| Tool call and MCP call routing rules | Trigger based on data source type | Differentiate access logic for ERP material data and on-site sensor data, matching multi-source data invocation scenarios |
| Document upload node supported formats | Calibrated based on actual testing | Cover commonly used formats for foundry categories, such as PDF process cards, Excel batching lists, CSV heat records, and others |
| Workflow trigger condition | Trigger based on heat batch | Match the update frequency of per heat and per batch for this category, ensuring data timeliness |
| Single recall count | Top 8 entries | Adapt to scenarios where per-batch data volume for foundry categories is limited, avoiding redundant data interference with quotation preparation |
nodeId binding | Bind the unique identifier of the corresponding data node | Accurately associate information from different sources such as melting furnace temperature measurement data and mold management system data |
Three Common Configuration Errors
- Issue: After configuring a form input node’s default value to reference
variablesvariables, the field appears empty or fails to load preset parameters. Root cause: The scope of thevariablesglobal variable is not properly configured, or the variable name has a case mismatch. - Issue: When the workflow calls data, both tool calls and MCP calls trigger simultaneously, causing a call conflict. Root cause: Routing rules are not configured based on data source type, and access permissions and invocation methods for different data sources are not differentiated.
- Issue: Uploaded foundry process cards cannot be recognized by the workflow. Root cause: Commonly used foundry document formats are not supported, or the upload node’s allowed format range is not correctly configured.
How to Verify Proper Configuration
- Trigger the workflow once, and confirm that the form input node’s default value automatically populates the preset
variablesglobal variables or form field values. - Simulate calls from different data source types, and confirm that only the matching data source type triggers the corresponding invocation method, with no conflicting calls.
- Upload commonly used PDF, Excel, and CSV format documents for foundry applications, and confirm that the workflow can correctly read the document content.
- View the logs for the node bound to
nodeId, and confirm that data from different sources is accurately associated with the corresponding workflow steps.
The values provided here are common starting points and should be evaluated against the reader’s own samples.