What Data for This Category Looks Like
Data for this category typically originates from silicon wafer incoming material records, manufacturing execution system data, real-time collection data from diffusion furnaces, coating equipment, and printers, EL and IV test results, and laboratory analysis data. Daily documents include process cards, diffusion and coating parameter tables, printing and sintering process records, EL inspection images, IV grading data, and yield analysis reports. Data update frequency adjusts dynamically based on production batches, single-wafer inspection results, and daily production summaries. Document fields cover texturing etching parameters, diffusion sheet resistance, coating thickness, printing line width, sintering temperature range, and other fields that change with the metrology standards of corresponding process links.
What Constraints These Characteristics Impose on the Tool Calling and Plugins Workflow
Data for this category is scattered and has flexible update rhythms. This requires tool calling to support real-time pulling and batch querying of multiple data sources, to avoid data lag affecting process parameter judgment. Document types are diverse and fields lack unified metrology standards. Plugins must adapt to multiple formats including process cards, inspection images, and numerical parameter tables, and require configuration of parameter extraction type verification rules. Single-wafer inspection data volume is small but batch updates are frequent. Tool calling context processing must support aggregating data by production batch, to avoid information overload from single queries. Process parameter accuracy directly relates to production yield. Tool calling results must be linked to original data traceability, to ensure Q&A content can be traced back to corresponding equipment or inspection records.
How to Set Configurations
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
| Workflow/API tool selection | auto | Adapt to randomly triggered multi-process tool calling requirements in process Q&A |
| Workflow/API function-call configuration | Enable parameter type verification | Ensure extracted process parameters fall within the field scope of photovoltaic cell manufacturing |
| Tool calling timeout setting | 300 seconds | Match the query time consumption of single-batch data, avoid failures caused by data pulling delays |
| Context aggregation cycle | 8–24 hours | Align with the data update rhythm of this category by production batch |
| Single-batch data query limit | Top 10 entries | Balance query efficiency and the completeness of single-batch data |
| Plugin call retry count | 3 times | Address temporary interruptions of equipment data, improve call success rate |
Three Common Mistakes
- Phenomenon: An invalid key error is returned when calling the knowledge base query interface, or the interface call returns a 401 status code. Cause: Correct application key permissions are not configured on the FastGPT platform, or credentials other than the key bound to the current plugin are used.
- Phenomenon: After the tool call is executed, the AI automatically appends unnecessary explanatory replies, and cannot return only the tool call result. Cause: The
toolChoiceparameter is not configured correctly, or the tool call result shielding function is not enabled. - Phenomenon: The parameters returned by the tool call do not match the actual process parameters, with field mismatches or unit errors. Cause: Parameter type verification for
functionCallis not enabled, and the exclusive field scope of photovoltaic cell manufacturing is not adapted.
How to Confirm the Configuration Is Complete
- Check whether the configurations of
toolChoiceandfunctionCallcomply with preset calling rules, and confirm that parameter verification is enabled. - Trigger a tool call for process parameter query, verify whether the returned result is linked to the production data of the corresponding batch, and there is no redundant non-tool call reply.
- Simulate a data pulling failure scenario, confirm that the plugin triggers the preset retry mechanism or returns a clear failure prompt.
- Confirm that the current FastGPT version is v4.17.0, to ensure that the plugin supports multi-data source association functions.
The values given are common starting points and should be measured against the reader's own samples.