### Answer Your AI agent is likely experiencing tool-calling issues where it announces an action but doesn't execute it. This is a known behavior pattern where LLMs struggle to determine when to call tools based solely on tool descriptions. [(1) ]() The core solution is to provide explicit tool-calling instructions in your prompt. LLMs often fail to call tools at appropriate times without clear triggers specified. For Single/Multi Prompt Agents, add explicit tool usage instructions that define exact conditions for when to call each tool:

///## Tool Usage Instructions
1. Gather initial information about the customer's issue.
2. Determine the type of request:
 - If customer mentions "refund" or "money back":
 → Call function `transfer_to_support` immediately
 - If customer needs order status:
 → Call function `check_order_status` with order_id
 - If customer wants to change their order:
 → First call `check_order_status`
 → Then transition to modification_state
3. After retrieving information:
 - Always summarize what you found
 - Ask if they need additional help
 - If yes, determine next appropriate action

Best practices for tool instructions: - Use trigger words: List specific words/phrases that should trigger tool calls - Define sequences: Specify when tools should be called in order - Set boundaries: Clarify when NOT to call certain tools - Reference tools by their exact names If tools are still not being triggered after updating your prompt, use the LLM Playground debug tools to identify what's going wrong, then apply fixes like adding conversation examples or adjusting LLM temperature. [(2) ]() [(3) ]() **

​ To learn more, see the sources I considered: [[1] Prompt Engineering Guide](https://docs.retellai.com/build/prompt-engineering-guide) [[2] Address Metric Issues](https://docs.retellai.com/ai-qa/address-metric-issues) [[3] Tools not being triggered](https://community.retellai.com/t/tools-not-being-triggered/201) ​

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

Thanks for the follow-up, @weronika. I'll get back to you soon.

### Answer I don't have information about the specific visual indicators (dashed blue lines, striped diagonal blue lines) in the Conversation Flow canvas based on the available sources. The documentation covers nodes, edges, and transition conditions in Conversation Flow [(1) ]() [(2) ]() , but doesn't explain the specific visual states you're describing (dashed outlines vs. striped patterns) . For clarification on these UI indicators, I recommend: * Visit [Documentation]() * Visit [Hire certified Retell partner]() Would you like a general answer based on my knowledge instead?

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

### Answer When your AI agent stops responding completely after a tool call, the most common cause is the "Speak After Execution" setting being disabled. [(1) ]() Solution: Enable "Speak After Execution" For tool/function nodes in your conversation flow, ensure the "Speak After Execution" option is enabled. When this setting is disabled, the agent executes the tool but doesn't generate a response afterward, causing it to appear "stuck" or unresponsive. Additional troubleshooting steps: If enabling "Speak After Execution" doesn't resolve the issue, check your function schema. [(2) ]() Incorrect schema configuration can prevent proper tool execution and response generation. For Conversation Flow agents experiencing stalling behavior: - Use the LLM Playground debug tools to identify what's going wrong [(3) ]() - Add conversation finetune examples if responses are inconsistent - Add transition finetune examples if the node isn't transitioning properly after tool execution - Consider splitting complex nodes into simpler, more deterministic steps - Adjust LLM temperature and use "Regenerate 10 Times" to test consistency If the agent continues to stall after these adjustments, the issue may involve node transition accuracy or other configuration problems that require deeper debugging through the LLM Playground. [(4) ]() [(3) ]()

​ To learn more, see the sources I considered: [[1] Tool calling - AI agent doesn't reply back. Stops interacting](https://community.retellai.com/t/tool-calling-ai-agent-doesnt-reply-back-stops-interacting/327) [[2] Conversation flow issue](https://discord.com/channels/1186424376300679209/1347631304790900736) [[3] Debug guide](https://docs.retellai.com/build/conversation-flow/debug-guide) [[4] Address Metric Issues](https://docs.retellai.com/ai-qa/address-metric-issues) ​

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] Update Conversation Flow](https://docs.retellai.com/api-references/update-conversation-flow) [[2] List Conversation Flows](https://docs.retellai.com/api-references/list-conversation-flows) [[3] Get Conversation Flow](https://docs.retellai.com/api-references/get-conversation-flow) ​

### Answer To set

high_priority

to

true

, you need to configure it within the

model_choice

object when creating or updating a conversation flow. [(1) ]() Here's how to structure it:

json
{
  "model_choice": {
    "type": "cascading",
    "model": "gpt-4.1",
    "high_priority": true
  }
}

[(2) ]() [(3) ]() [(1) ]() The

high_priority

field is a boolean property within the

model_choice

configuration. When set to

true

, it prioritizes your agent's requests. This applies when using the API to create or update conversation flows - you'll include this in your request body under the

model_choice

parameter. [(1) ]()

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