Data profile for this category
Data sources for this category include the national land spatial planning unified map, project approval and supervision platform, current land use database, and ecological protection red line data. Update frequency is not fixed on a schedule. Documents routinely produced and used include land use preliminary examination and site selection opinions, construction land planning permits, construction engineering planning permits, vector red line maps, and planning condition documents. Core fields cover project name, land use location, land use scale, planning control requirements, pre-approval node status, and more. There is no fixed unified format template. Some documents use vector format, while others use PDF or structured table formats.
Constraints on tool calling and plugin workflows
Scattered data sources and diverse formats require tool calling to support pulling and parsing multi-source heterogeneous data, and to handle documents in formats including vector maps, PDFs, and structured tables. Since data updates do not follow a fixed cycle, tool calling must support pulling the latest data on demand, to avoid outdated data caused by fixed caching. Business errors directly affect project approval and construction startup, so tool calling workflows must include pre-check logic to ensure pulled data complies with planning control requirements. Additionally, the business requires coordination of multi-department preconditions, so tool calling must support linked calls across platform interfaces.
Configuration settings
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
toolChoice | auto | Adapts to multi-type tool calling needs, covering data pulling, document parsing, conflict checking, and other links |
functionCall | Enabled | Supports outputting ledger content according to preset structured fields, facilitating automated summary for subsequent report generation |
| Plugin call timeout | 120 seconds | Requires pulling multi-source department data such as the national land spatial planning unified map, to avoid task interruption caused by interface delays |
| Number of parallel multi-source data calls | First 3 requests | Prevents excessive concurrency from exceeding interface call limits of natural resource department platforms |
| Content check switch | Enabled | Requires checking conflicts between land red lines and ecological protection red lines, to ensure output results comply with control requirements |
| Document parsing format | Auto-adapt | Covers multiple document types including vector red line maps, PDF certificates, and structured tables |
Three common misconfigurations
- Symptom: A 504 timeout error is returned when the workflow calls external department interfaces. Cause: The timeout configuration was not adjusted for interface delays of natural resource department platforms, and the default short timeout duration was used.
Reference only; verify against the target FastGPT release and deployment documentation.
- Symptom: Streamed output contains unauthorized land use prompts that are not filtered. Cause: The content check plugin was not configured, and the output content filtering link was not enabled.
How to confirm correct configuration
- Initiate a single test call to pull approval data for a single project, and check whether the returned results include core fields such as project name, land use location, and land use scale.
- View workflow logs to confirm there are no timeout or connection failure errors in the multi-source data pulling link, and that plugin components load normally.
- Trigger the report generation process, and check whether the output ledger is displayed in structured format according to preset fields, with no missing or disorganized content.
- Simulate 2 to 3 concurrent calls, and observe whether the workflow can process requests stably without abnormal exits or errors.
The values provided here are common starting points, and should be evaluated against the user'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.