What the Data for This Category Looks Like
Daily documents for campaign setup include campaign plans, targeting settings, budget scheduling, and conversion tracking configuration. Data sources include advertising platform APIs, user profile systems, budget spreadsheets, and attribution tools. Update frequency is daily or per individual campaign cycle. Document fields include targeting dimension parameters, budget amount and unit, conversion event type, attribution rule configuration, and more. Each field must strictly match the format requirements of the corresponding data source, with no unified fixed format.
Constraints Imposed on Tool Calling and Plugins
Multi-source data dependencies require tool calling to adapt to the API formats of different advertising platforms. Use plugins to complete cross-data source field mapping. As a pre-launch advertising configuration step, high-frequency data updates require tool calling cache durations to match data update rhythms. This prevents use of expired campaign configuration data. Multi-field structures require explicit validation of returned field legitimacy during tool calling. This prevents campaign plans from failing to launch due to field mismatches. Binding between documents and data source formats requires tool calling parameters to strictly follow the naming rules of the corresponding platform. This avoids data misalignment.
Configuration Settings
| Configuration Item | Recommended Value | Rationale |
|---|---|---|
toolChoice | auto | Automatically select tools adapted to different advertising platforms, no manual tool specification required |
functionCall | Mandatorily return conversion tracking configuration fields | Ensure fields returned by the tool match the conversion tracking configuration in campaign setup documents |
| Tool calling cache duration | 300 seconds | Match the daily update frequency of campaign plans, avoid using expired campaign configuration data |
| Call timeout threshold | 60 seconds | Prevent workflow interruptions caused by delayed responses from advertising platform APIs |
| Single call result limit | 10 entries | Match the number of fields in campaign setup documents, avoid returning excessive redundant data |
| Context window size | 2000 characters | Adapt to the text length requirements of campaign setup documents |
Three Common Mistakes
- Unable to retrieve campaign plan parameter values returned by MCP tools in workflows. The symptom is empty node output fields. The cause is failure to explicitly specify the return field range in the
functionCallconfiguration. - When calling the MongoDB query tool, token consumption reported by the model does not match actual API call consumption. The symptom is differing token values in logs. The cause is default inclusion of redundant context metadata during tool calling.
- Irrelevant campaign performance data is returned after knowledge base calls. The symptom is response content deviating from the current campaign setup query topic. The cause is failure to apply targeted restrictions to the tool calling data source scope.
How to Confirm Configurations Are Correct
- Start a test workflow, check the output logs of the tool calling node, confirm returned fields match the range specified in the configuration.
- Compare token consumption reported by the model with actual consumption records from MongoDB or advertising platform APIs, confirm the difference is within expected ranges.
- Input queries related to campaign setup, verify returned content only includes information from the currently configured data sources.
- Adjust the tool calling timeout threshold in the FastGPT v4.17.0 environment, test workflow response when the threshold is exceeded, confirm exception handling logic takes effect.
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 2026-10-19.