What the data for this category looks like
Data sources include mine production scheduling systems, mineral processing equipment collection terminals, and geological survey and exploration archive databases. Update frequency follows two schedules: real-time production parameters are synchronized every hour, and exploration archives are updated every quarter. Document structure includes stope unique identifier, ore type, mineral processing process parameters, equipment operating duration, and safety inspection identifier. Field units include hours, tons, and cubic meters. Percentage-based statistical values are not included.
What constraints these characteristics impose on the tool calling and plugins link
Real-time production parameters synchronized every hour require setting reasonable request intervals for tool calls to avoid interface current limiting caused by frequent pulls. Exploration archives updated every quarter mean there is no need to refresh archive data frequently when calling plugins, and a fixed cache cycle can be configured. Document fields include exclusive identifiers such as stope unique identifier and mineral processing process parameters. Input parameters for tool calls must strictly match these field names; general fields cannot be used directly for data association. Equipment operating duration is measured in hours. Results returned by plugins must use a unified unit format to avoid process guidance errors caused by unit mismatches.
How to Set Configurations
Reference only; verify against the target FastGPT release and deployment documentation.
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
toolChoice | auto | Adapt to the variable process guidance needs in mining and mineral processing business, automatically match available tools |
functionCall | enabled | Support calling exclusive tools for mine production and safety to complete process node verification |
| Tool request interval | 3600 seconds | Match the update rhythm of real-time production parameters synchronized every hour, avoid current limiting caused by frequent requests |
| Archive data cache duration | 7 days | Exploration archives are updated every quarter. A 7-day cache balances data timeliness and request efficiency |
| Tool input parameter verification rule | Mandatorily match stope unique identifier field | Document fields include exclusive identifiers, strict verification is required to ensure accurate data association |
| Number of entries returned per tool call | Top 3 entries | Business process nodes for mining and mineral processing are relatively fixed, no excessive redundant data is needed |
Three Common Misconfigurations
- Phenomenon: Returns an error message starting with
InternalError.Algo.InvalidParameter: The tool. Cause: Tool input parameters do not match the exclusive fields for mining and mineral processing, such as failing to pass the stope unique identifier, resulting in parameter verification failure. - Phenomenon: Plugin calls return an error message containing "Failed to get the VQD for query". Cause: Tool request intervals are not set reasonably, frequent pulling of real-time production data triggers interface current limiting, resulting in failed data acquisition.
- Phenomenon: Missing mineral processing process parameters in process guidance. Cause: The
functionCallconfiguration is not enabled, so exclusive tools cannot be called to complete automatic verification and completion of process parameters.
How to Confirm the Configuration Is Properly Set
- Initiate a test request that includes the stope unique identifier, check whether the parameters returned by the tool call match the fields defined in the document.
- View tool call logs to confirm that the request interval conforms to the preset real-time data synchronization cycle.
- Trigger archive data calls to confirm that the update cycle of cached data conforms to the update rhythm of business archives.
- Check the configuration status of
toolChoiceandfunctionCallto confirm that both functions are enabled.
The values given are common starting points and should be measured against the reader's own samples.
Question material comes from public community discussions. Configuration values are common starting points and should be measured against your own samples. Verified on 2026-09-28.