What Data for This Category Looks Like
This category consists of structured public data assets. Sources include local public data open platforms, integrated big data platforms, departmental business databases, and electronic credential repositories. Supporting documentation for routine maintenance includes public data resource catalogs, data opening lists, data item descriptions, desensitization rules, interface documentation, and update frequency explanations. Data fields cover data item identifiers, affiliated departments, opening scopes, desensitization methods, interface addresses, update cycles, and more. The size of a single record ranges from tens of KB to several MB. The overall update frequency is monthly.
Constraints on Tool Calling and Plugins From These Characteristics
Public data catalogs have structured characteristics. Tool calls must strictly match preset fields to avoid returning unauthorized data. The overall update frequency is monthly. Tool calls must refresh cached data regularly to avoid returning expired or redundant data. Data formats vary across multiple sources. Plugins must adapt to interface specifications of different data sources and unify data return formats. Errors such as data leaks and inconsistencies between catalogs and actual data can have serious consequences. Tool calls must verify authorization scopes and enable data desensitization configurations. This ensures only content meeting opening conditions is returned. Service consultation scenarios require concise responses. Tool call results must directly return structured data without additional spliced content.
How to Configure Settings
Reference only; verify against the target FastGPT release and deployment documentation.
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
toolChoice | auto | Adapt to multi-type calling requirements of public data catalogs, automatically select appropriate tools or functions |
functionCall | Enabled | Used to call preset public data query functions to obtain structured catalog data |
| Tool call timeout | 600 seconds | Public data interfaces may respond slowly; prevents early termination of requests |
| Disable knowledge base input/output citations | Enabled | Only returns structured data results after tool calls, meeting the concise response requirements of service consultation |
| Bind unique identifier to workflow | Bind according to business scenarios | Distinguish public data catalog calls from different departments, avoiding cross-department data confusion |
Three Common Misconfigurations
- Tool calls return unsensitized sensitive data. Cause: Data desensitization configuration is not enabled, and uncleaned departmental business database data is called directly, violating security requirements for data opening.
- Workflow calls return 401 errors. Cause: A globally universal API Key is used to call department-specific public data interfaces, and the corresponding department's exclusive authorization key is not used.
- Knowledge base search citation fragments appear in responses. Cause: The knowledge base input/output citation configuration item is not disabled, causing tool call results to be contaminated by additional knowledge base content.
How to Verify Successful Configuration
- Initiate a service consultation request, confirm the returned result only contains public data content obtained via tool calls, with no additional knowledge base citation fragments.
- View the binding identifier of the workflow, confirm each department's public data call uses an exclusive unique ID to avoid data confusion.
- After triggering a tool call, confirm the fields of the returned result completely match the definitions in the public data item descriptions, with no missing or formatting errors.
- Simulate a timeout scenario, confirm the request is not terminated early when no response is received after more than 600 seconds.
The values provided 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.