​ To learn more, see the sources I considered: [[1] List Calls](https://docs.retellai.com/api-references/list-calls) [[2] Webhook Overview](https://docs.retellai.com/features/webhook-overview) ​

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

### Answer To find calls with transfer failures, you can use the List Calls API with filter criteria to identify calls where transfers failed [(1) ]() . The List Calls API supports filtering by various call attributes . You can filter calls using the

filter_criteria

object in your API request . Here's how to structure your request:

javascript
import Retell from 'retell-sdk';
const client = new Retell({
 apiKey: 'YOUR_RETELL_API_KEY',
}) ;
const callResponses = await client.call.list() ;
console.log(callResponses) ;

The response will include a

disconnection_reason

field for each call . Transfer-related disconnection reasons you should look for include

call_transfer

. Additionally, webhook events can help you track transfer failures in real-time [(2) ]() . Retell AI supports these transfer-related webhook events : -

transfer_started

- Triggered when a transfer is initiated -

transfer_bridged

- Triggered when a transfer successfully bridges -

transfer_cancelled

- Triggered when a transfer is cancelled or fails to connect -

transfer_ended

- Triggered when the transfer leg ends The webhook payload will include the

transfer_destination

field when applicable [(2) ]() [(1) ]() . To specifically identify failed transfers, look for calls with

transfer_cancelled

webhook events or check the

disconnection_reason

field in your call data .

### Answer Based on the available sources, I don't have specific information about finding transfer failure calls directly in the Retell Dashboard interface. The sources primarily document the API methods for retrieving call information [(1) ]() [(2) ]() . However, you can check the

disconnection_reason

field through the Retell Dashboard to diagnose specific failure reasons [(3) ]() . The dashboard should display call details including disconnection reasons like

call_transfer

[(1) ]() [(2) ]() . For programmatic access, you can use the List Calls API to filter and retrieve calls with transfer-related issues [(1) ]() . The API response includes the

disconnection_reason

field that indicates why calls failed [(1) ]() [(2) ]() . If you need more specific guidance on navigating the dashboard to find these calls, I recommend: - Visit [Documentation]() - Visit [Hire certified Retell partner]() Would you like a general answer based on my knowledge instead?

​ To learn more, see the sources I considered: [[1] List Calls](https://docs.retellai.com/api-references/list-calls) [[2] Get Call](https://docs.retellai.com/api-references/get-call) [[3] frequent outbound call fail during transfer](https://community.retellai.com/t/frequent-outbound-call-fail-during-transfer/774) ​

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

Thanks for the follow-up, @Q4. I'll get back to you soon.

​ To learn more, see the sources I considered: [[1] Get Voice Agent](https://docs.retellai.com/api-references/get-agent) [[2] List Voice Agents](https://docs.retellai.com/api-references/list-agents) [[3] Debug call disconnection](https://docs.retellai.com/reliability/debug-call-disconnect) ​

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

### Answer Error ASR (

error_asr

) is a disconnection reason indicating that Retell's Automatic Speech Recognition (ASR) encountered a problem during the call [(1) ]() . When this error occurs, the call status is marked as

error

. ASR is the technology that converts spoken language into written text [(2) ]() . It's the first critical step in enabling AI voice agents to understand and respond to callers . When the ASR system fails, the agent cannot accurately transcribe what the caller is saying, which prevents the conversation from continuing . This is considered an error condition rather than expected behavior [(1) ]() . If you're experiencing

error_asr

disconnection reasons, it indicates a technical issue with the speech recognition processing that prevented the call from completing successfully .

​ To learn more, see the sources I considered: [[1] Debug call disconnection](https://docs.retellai.com/reliability/debug-call-disconnect) [[2] Automatic Speech Recognition (ASR)](https://www.retellai.com/glossary/automatic-speech-recognition-asr) ​

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

### Answer These are disconnection reasons that explain why calls end or fail to connect [(1) ]() . Here's what each one means: Expected Behavior (Normal Call Endings) : - **User Hangup**: User hung up the call - **Agent Hangup**: AI agent hung up the call - **Call Transfer**: AI agent transferred the call - **Voicemail Reached**: Voicemail is reached (when agent is configured with voicemail settings) - **IVR Reached**: IVR system is reached (when agent is configured to hang up upon encountering IVR) - **Inactivity**: Call terminated due to "end_call_after_silence_ms" setting after long inactivity - **Max Duration Reached**: Call terminated due to maximum duration reached Outbound Call Not Connected: - **Dial Busy**: The number dialed is busy - **Dial Failed**: Dialing failed with no or unknown SIP error code - **Dial No Answer**: The number dialed did not answer - **Invalid Destination**: The number dialed is invalid - **Telephony Provider Permission Denied**: SIP trunk credentials are not authenticated - **Telephony Provider Unavailable**: The telephony provider is unavailable - **SIP Routing Error**: SIP routing is going over too many hops or is in a loop - **Marked As Spam**: The number dialed is marked as spam - **User Declined**: User declined the call Error Conditions: - **Concurrency Limit Reached**: Concurrency limit reached, add retry with exponential backoff or consider enterprise plan - **No Valid Payment**: No valid payment registered on file, or service shut down due to bill overdue - **Error LLM Websocket Open**: LLM websocket did not open between Retell server and your backend (likely incorrect Custom LLM URL or unreachable LLM server) - **Error LLM Websocket Lost Connection**: LLM websocket connection broke during the call - *Error LLM Websocket Runtime

*: LLM websocket received a closing signal other than

1000

from your server - **Error LLM Websocket Corrupt Payload**: LLM websocket received unspecified payload - **Error No Audio Received**: Has not received audio from Twilio or web frontend for a while after connection has established - **Error ASR**: Retell's ASR encountered a problem - **Error Retell**: Unspecified Retell side problem - **Error Unknown**: Unknown error - **Error User Not Joined**: User did not join web call within 30s after calling startWebCall - **Registered Call Timeout**: Phone call is 30s or more apart from registration Transfer-Related: - **Transfer Bridged**: Not listed in the standard disconnection reasons, but appears in webhook events when a transfer successfully bridges - **Transfer Cancelled**: Not listed in the standard disconnection reasons, but appears in webhook events when a transfer is cancelled or fails to connect You can check these disconnection reasons through the Retell Dashboard or by using the get-call API endpoint .

​ To learn more, see the sources I considered: [[1] Debug call disconnection](https://docs.retellai.com/reliability/debug-call-disconnect) ​

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

### Answer Yes, you can create a complete voice agent using just the API [(1) ]() . To create a voice agent via API, you need to provide a response engine and voice configuration . Here's an example:

javascript
import Retell from 'retell-sdk';
const client = new Retell({
 apiKey: 'YOUR_RETELL_API_KEY',
}) ;
const agentResponse = await client.agent.create({
 response_engine: { llm_id: 'llm_234sdertfsdsfsdf', type: 'retell-llm' },
 voice_id: '11labs-Adrian',
}) ;
console.log(agentResponse.agent_id) ;

The Create Voice Agent endpoint requires: - **response_engine**: The Response Engine to attach to the agent (you need to create a Response Engine first) - **voice_id**: Unique voice id used for the agent You can also configure optional parameters like agent_name, voice_model, responsiveness, interruption_sensitivity, webhook_url, and many other settings . Make sure your voice_id and llm_id are correct and valid [(2) ]() . If you use an incorrect or obsolete voice ID, you may receive a 404 error .

​ To learn more, see the sources I considered: [[1] Create Voice Agent](https://docs.retellai.com/api-references/create-agent) [[2] Create Agent (API endpoint)](https://discord.com/channels/1186424376300679209/1424438362458161182) ​

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

### Answer Yes, you can create a fully configured, working voice agent using just the API [(1) ]() . The Create Voice Agent endpoint supports all configuration options available in the dashboard . Here's a complete example with comprehensive configuration:

javascript
import Retell from 'retell-sdk';
const client = new Retell({
 apiKey: 'YOUR_RETELL_API_KEY',
}) ;
const agentResponse = await client.agent.create({
 response_engine: { llm_id: 'llm_234sdertfsdsfsdf', type: 'retell-llm' },
 voice_id: '11labs-Adrian',
}) ;
console.log(agentResponse.agent_id) ;

You can include extensive optional parameters to fully configure your agent : - **agent_name**: Name for your reference - **voice_model**: Select specific voice model (e.g., "eleven_turbo_v2") - voice_speed, voice_temperature, **volume**: Control voice characteristics - responsiveness, **interruption_sensitivity**: Control conversation dynamics - enable_backchannel, **backchannel_frequency**: Configure backchanneling behavior - **language**: Specify language/dialect (e.g., "en-US") - webhook_url, **webhook_events**: Set up event notifications - enable_voicemail_detection, **voicemail_message**: Configure voicemail handling - **ivr_option**: Define IVR detection behavior - end_call_after_silence_ms, **max_call_duration_ms**: Set call timeout rules - **post_call_analysis_data**: Define data extraction after calls The only prerequisites are creating a Response Engine first (like a Retell LLM) and having a valid voice_id . Once created, the agent is immediately functional and can handle calls .

​ To learn more, see the sources I considered: [[1] Create Voice Agent](https://docs.retellai.com/api-references/create-agent) ​

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

### Answer When creating an agent via API, the prompt configuration depends on the response engine type you attach [(1) ]() [(2) ]() . For Single/Multi Prompt Agents: The prompt is configured in the Response Engine (Retell LLM) , not directly in the agent creation [(3) ]() [(4) ]() . You first create a Retell LLM with your prompt, then attach it to the agent using the

response_engine

parameter [(1) ]() [(2) ]() . Agent Type Determination: The API creates a Single Prompt Agent by default when you use a Retell LLM as the response engine [(3) ]() [(5) ]() . The single prompt approach uses one comprehensive prompt to define the agent's behavior [(5) ]() . Multi-Prompt Agents: Multi-prompt combines several single prompt agents into a structured tree of prompts, where each node can contain its own specific prompt and transition logic . Conversation Flow Agents: Conversation Flow provides the most control over dialogue paths and is designed for users who prefer not to engage in extensive prompt engineering . It uses a node-based visual editor where each node focuses on one specific task . To create different agent types, you would need to configure the appropriate response engine structure. The basic API example shown creates a single prompt agent by attaching a Retell

LLM [(1) ]() [(2) ]() . For more details on configuring prompts and choosing between agent types, check out: - [Prompt Engineering Guide]() - [Step 2: Configure the basic settings]() **

​ To learn more, see the sources I considered: [[1] Get Voice Agent](https://docs.retellai.com/api-references/get-agent) [[2] Update Voice Agent](https://docs.retellai.com/api-references/update-agent) [[3] Prompt Engineering Guide](https://docs.retellai.com/build/prompt-engineering-guide) [[4] Step 2: Configure the basic settings](https://docs.retellai.com/build/single-multi-prompt/configure-basic-settings) [[5] Retell AI’s Advanced Conversation Flow | Retell AI](https://www.retellai.com/blog/unlocking-complex-interactions-with-retell-ais-conversation-flow) ​

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

### Answer When you create or retrieve a shared (library) component, the API response includes a

linked_conversation_flow_ids

field that lists all conversation flow IDs where that component is being used [(1) ]() . Here's what the API returns:

json
{
  "conversation_flow_component_id": "<string>",
  "name": "Customer Information Collector",
  "linked_conversation_flow_ids": [
    "<string>"
  ]
}

The

linked_conversation_flow_ids

array contains the IDs of all conversation flows that reference this shared component . If your component is used in 2 conversation flows, you'll see 2 flow IDs in this array . In the Dashboard: When viewing a shared component in the Components tab under Library Components, you can see which agents use it [(2) ]() . The component tracks all conversation flows it's linked to . Important behavior: When you delete a shared component from the library, linked instances in agents are downgraded to local copies and stop receiving sync updates, but your agents keep working .

​ To learn more, see the sources I considered: [[1] Create Conversation Flow Component](https://docs.retellai.com/api-references/create-conversation-flow-component) [[2] Components](https://docs.retellai.com/build/conversation-flow/components) ​

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

### Answer To identify which agents are using a specific component, you need to map from conversation flows to agents [(1) ]() . Step 1: Get the conversation flow IDs When you retrieve a shared component via API, the response includes

linked_conversation_flow_ids

that shows which conversation flows use that component [(2) ]() :

javascript
{
  "conversation_flow_component_id": "<string>",
  "linked_conversation_flow_ids": [
    "<string>"
  ]
}

Step 2: Match conversation flows to agents Once you have the conversation flow IDs, you need to identify which agents use those conversation flows [(1) ]() . Each agent has a

response_engine

that references a specific conversation flow . When you create an agent, you attach a conversation flow ID as the response engine :

javascript
{
  "response_engine": { 
    "conversation_flow_id": "llm_234sdertfsdsfsdf", 
    "type": "conversation-flow" 
  }
}

To find all agents using a specific conversation flow (and therefore your component) , you would need to list your agents and check which ones have that conversation flow ID in their response engine configuration . In the Dashboard: The dashboard shows you which conversation flows use a component when you view it in the Components tab under Library Components [(3) ]() . However, to see the actual agent names/IDs, you'd need to cross-reference those conversation flow IDs with your agents [(1) ]() .

​ To learn more, see the sources I considered: [[1] Discord message from @andrii09152 on 2025-09-11](https://discord.com/channels/1186424376300679209/1186462578310910112/1415820290251952189) [[2] List Conversation Flow Components](https://docs.retellai.com/api-references/list-conversation-flow-components) [[3] Components](https://docs.retellai.com/build/conversation-flow/components) ​

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