When calling a custom tool from a conversation flo...
# ❓・ai-evy-supportw
weronika
04/14/2026, 12:54 PM@AI Evy When calling a custom tool from a conversation flow node, what's the best practice for where to specify the parameter values?
Should I:
(a) Put all parameter values and logic in the node prompt (e.g. "Call get_availability with: is_new_patient: false, is_emergency: false, appointment_type: [what the patient asked for]...")
(b) Rely on the tool's parameter descriptions (in the tool schema) to guide the model, and keep node prompts short?
(c) Do both — repeat key context-specific values in the node prompt even when they're also described in the tool schema?
a
AI Evy
04/14/2026, 12:54 PMThanks for tagging me @weronika. I'll look through relevant sources and get back to you shortly.
AI Evy
04/14/2026, 12:55 PM### Answer
I don't have specific documentation about best practices for specifying parameter values when calling custom tools from conversation flow nodes.
However, based on the available documentation, custom functions use JSON schema to define parameters [(1) ]() , and the LLM uses these parameter descriptions to determine what arguments to pass . The parameter schema should include clear descriptions for each field .
For more specific guidance on this architectural decision, I'd suggest:
Check out these potentially useful sources:
- [Custom Function]()
- [Prompt Engineering Guide]()
- [Create Conversation Flow]()
You might also want to reach out for more detailed guidance:
- Visit [Documentation]()
- Visit [Hire certified Retell partner]()
AI Evy
04/14/2026, 12:55 PM
To learn more, see the sources I considered:
[[1] Custom Function](https://docs.retellai.com/build/conversation-flow/custom-function)
AI Evy
04/14/2026, 12:55 PMFeel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.