### Answer --- 1. Built-in System Variable for Current Time/Day Yes, Retell provides a default system variable

{{current_time}}

that includes the current day of the week. [(1) ]() The format is:

"Thursday, March 28, 2024 at 11:46:04 PM PST"

You can also use timezone-specific variants like

{{current_time_America/New_York}}

. You can reference

{{current_time}}

directly in your System Prompt to enable the LLM to determine the day and apply your routing logic. --- 2. Inbound Webhook for Dynamic Variables If you prefer to calculate the day server-side and pass a custom variable like

is_office_open

, you can use the Inbound Call Webhook. [(2) ]() Your webhook receives a POST with the inbound call details, and you respond with

dynamic_variables

under the

call_inbound

key. Example webhook response:

json
{
  "call_inbound": {
    "dynamic_variables": {
      "is_office_open": "false",
      "current_day": "Wednesday"
    }
  }
}

Your server-side logic would determine the day and set

is_office_open

accordingly before returning this response. --- Prompt Tip: In your System Prompt, you can then include logic like: > "If

{{is_office_open}}

is false, do not use the transfer_call tool. Instead, offer to schedule a callback." Or if using the built-in variable: > "Check

{{current_time}}

for the current day. If it is Wednesday or Friday, do not transfer—offer a callback instead." ---

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.

​ To learn more, see the sources I considered: [[1] Dynamic Variables](https://docs.retellai.com/build/dynamic-variables) [[2] Inbound webhook](https://docs.retellai.com/features/inbound-call-webhook) ​

### Answer --- To ensure the agent knows the current time/day during the call, use the

{{current_time}}

system variable directly in your prompt. [(1) ]() How it works: The

{{current_time}}

variable is a default system variable that Retell automatically provides—it outputs the current time including the day of the week. Example output:

"Thursday, March 28, 2024 at 11:46:04 PM PST"

You can also specify a timezone: -

{{current_time_America/New_York}}

for Eastern time -

{{current_time_Australia/Sydney}}

for Sydney time To verify the agent is using it correctly: 1. Include it explicitly in your prompt, for example: > "The current time is {{current_time}}. If the day is Wednesday or Friday, do not use the transfer_call tool—offer a callback instead." 2. Test in the LLM Playground before deploying—you can see how the variable resolves and how the agent responds to different scenarios. 3. Review call transcripts after test calls to confirm the agent referenced the correct day/time in its decision-making. The variable is evaluated at runtime, so the agent will always have access to the real current time during the conversation. ---

​ To learn more, see the sources I considered: [[1] Dynamic Variables](https://docs.retellai.com/build/dynamic-variables) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.

### Answer --- Based on the documentation, here's what I can address: 1. Is

contains

the correct operator? Yes,

CONTAINS

is a valid operator for equation-based transitions. The documentation shows examples like: [(1) ]()

"New York, Los Angeles" CONTAINS {{user_location}}

However, note that the

CONTAINS

operator performs string comparisons. 2. How to debug the runtime value: You can view node transitions in the call transcript within the History tab—it shows the node names the agent transitions from and to. Additionally, you can track how a past call transitioned through each node in the conversation flow using "View Call History in LLM Playground." [(2) ]() 3. Why Friday might match Transition 1: The documentation states that all equation conditions are evaluated from top to bottom, and the agent travels on the first condition that evaluates to true. [(1) ]() The issue is likely the order of evaluation within your OR conditions. Since Transition 1 is evaluated first and contains multiple OR conditions, if any of those conditions unexpectedly evaluates to true, it will take that path before ever checking Transition 2. Potential causes: - There may be an issue with how the

{{current_time_Europe/Amsterdam}}

variable is being resolved at runtime - Community users have reported similar issues where equation-based transitions with dynamic variables don't evaluate as expected [(3) ]() Recommended debugging steps: 1. Check the call transcript in the History tab to see which nodes were traversed [(1) ]() 2. Consider using the LLM Playground to test how variables resolve [(4) ]() 3. Try adding a conversation node before the split that outputs the value of

{{current_time_Europe/Amsterdam}}

to verify what the agent actually sees Alternative approach: If equation-based transitions continue to be unrelia

ble, you could use prompt-based transition conditions instead, which let the LLM evaluate the logic. Example: [(1) ]()

User said something about booking a meeting

Or calculate the day server-side via an Inbound Webhook and pass a simpler boolean like

is_office_open

as a dynamic variable, then use:

{{is_office_open}} == "true"

---

​ To learn more, see the sources I considered: [[1] Step 3: Add transition conditions](https://docs.retellai.com/build/conversation-flow/transition-condition) [[2] Platform Changelogs | Retell AI](https://www.retellai.com/changelog) [[3] Transitions do not work with equations and dynamic variable](https://discord.com/channels/1186424376300679209/1430615179967856663) [[4] Debug guide](https://docs.retellai.com/build/conversation-flow/debug-guide) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.

### Answer --- The issue you're experiencing—where the agent speaks the instruction text instead of remaining silent—is likely related to how Conversation Nodes work. When inside a Conversation Node, the agent will generate content based on the instruction, even if you tell it to "say nothing." [(1) ]() Why this happens: Conversation Nodes are designed for having conversations with users. The agent will attempt to generate a response based on the instruction you provide. Writing "Say NOTHING" as an instruction doesn't prevent the agent from speaking—it just gives the LLM that text to interpret, which can lead to unexpected verbalization. Recommended solution: Use the "Skip Response" setting Enable Skip Response on your

HUMAN_TRANSFER_NODE

. When this setting is enabled, the transition will only have one edge you can connect, and when the agent is done (in this case, immediately since there's nothing to say) , it will transition to the next node via that specific edge. However, note that Skip Response only allows one outgoing edge, which conflicts with your need for 3 conditional transitions. Alternative approach: Restructure your flow Since you need multiple equation-based transitions, consider: 1. Use a Function Node instead for silent routing logic—Function Nodes can execute without the agent speaking during execution. [(2) ]() 2. Pre-evaluate the routing before reaching this node. Have your Make.com webhook return a single variable like

route_destination

with values

"transfer"

,

"callback"

, or

"outside_hours"

, then use separate paths earlier in your flow. Regarding dynamic variable evaluation: For equation conditions, you can only use variables that are passed in as dynamic variables.

[(3) ]() Verify your webhook response is being received correctly by checking the call transcript in the History tab—it shows node transitions so you can see the actual path taken. Debug steps: - Review call transcripts in the History tab to see which nodes were traversed - Add a test conversation node that explicitly speaks the value of

{{is_office_open}}

to confirm the variable is being set correctly - Consider using the LLM Playground to test variable resolution [(2) ]()

​ To learn more, see the sources I considered: [[1] Conversation Node](https://docs.retellai.com/build/conversation-flow/conversation-node) [[2] Debug guide](https://docs.retellai.com/build/conversation-flow/debug-guide) [[3] Step 3: Add transition conditions](https://docs.retellai.com/build/conversation-flow/transition-condition) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.