Skip to content
EN
Open app

For AI assistants: this page as plain markdown is at https://docs.voicelogica.ai/raw/agent-configuration/built-in-tools.md, and https://docs.voicelogica.ai/llms.txt lists the docs by topic, one file each, small enough to read whole. Read those rather than this HTML page.

Tools & Workflows

11 min read
Copy as Markdown
View as MarkdownOpen the raw .md file.md

Skills - Tools & Workflows is the permission list: what this agent is allowed to do, not what it knows. A tool that is off cannot fire, even if the prompt asks for it. A workflow that is not attached here cannot run, even if you built it under Workflows.

The header shows how many tools are enabled (for example 6 of 51 tools enabled) and how many integrations are connected.

Tools and Workflows overview
toolsExpand

Do not enable the whole catalog on day one. Extra tools are extra ways to go off-script. The onboarding agent already has a small Enabled set - that is enough for rehearsal.

A multi-step job you already built (API, email, branch)
TURN ONBusiness WorkflowsNEEDSAttach it here or the agent cannot run it. Empty state shows No workflows added yet. Use Add workflow to create and link one.
One job the product already ships (SMS, skip-turn, voicemail)
TURN ONBuilt-in ToolsNEEDSToggle + gear. Enabled is the truth; browse the catalog with the other filters.
A remote tool server (Claude / a custom MCP)
TURN ONMCP serversNEEDSAdd MCP, connect (Streamable HTTP, Bearer, or OAuth), then pick which of its tools this agent may call.

First setup - a receptionist that does not talk over digits

Section titled “First setup - a receptionist that does not talk over digits”
  1. Open Skills - Tools & Workflows

    Note N of M tools enabled. That number is what the model is allowed to call.

  2. Leave Business Workflows empty until you have a real job

    Then Add workflow or attach an existing row and turn its toggle on. The pencil opens the workflow editor.

  3. On Built-in Tools, start from Enabled

    Filters include Enabled, Essentials, All categories, Conversation, Outbound, and Integrations. Enabled is what this agent will actually use. Use the other filters to find something to turn on.

  4. Keep these on for a first inbound line

    A typical onboarding set includes:

    Skip Turn
    The caller is about to read a number or code. Without it the agent talks over the digits. Open the gear (Skip Turn - Tool Settings). Use a template: Phone number, Product code, Order number, or Continue after Input. Timeout and max skips per turn control how long it stays silent.
    Handle Silence
    The line goes quiet. Configure the interval, how many nudges, then Forward the call and/or End the call. If the action is forward/transfer, Transfer under Call handling must be ready or Required can block you.
    Detect Voice Mail
    Outbound or missed-call callbacks, so you do not “talk” to a mailbox.
    Lock on Speaker
    Helps the agent stay locked on the right speaker when helpful for your line.
    Skip Turn tool settings with phone number template
    tools-skipExpand
  5. Add SMS only if you will actually text

    Send SMS and Capture Data via SMS use credits and must be asked for. Put one line in the personality prompt (“Text a confirmation when they ask”) so the model knows it is allowed - the toggle alone is not a brief.

    Send Info Page by SMS writes the information the caller asked for (prices, steps, links) as a web page and texts them the link. Opens and link clicks show under Tracked links on the call. On calls a person answers, turn it on on the analysis agent so the wake-word assistant can send what knowledge assist found.

  6. Play wait music on tools

    When on, callers hear music instead of silence while a tool runs. Leave it off if the agent mostly talks without pausing for tools.

  7. Try this agent

    Read a phone number slowly. If the agent talks over you, Skip Turn is off or its prompt is vague. Go quiet for a bit: Handle Silence should nudge, then forward or end as configured.

Leave CRM, IVR, appointments, and sentiment off until that job exists. Tools that show Connect / Authorize stay blocked until an admin connects them under Integrations (or the button on the row).

Transfer itself is configured under Call handling - Transfer (Warm is often recommended; Attended and Blind are also available) - enable the tool path here only when that job needs it.

The catalog is long, but these come up often. Turn one on only when its job exists, and give the model one line in the personality prompt so it knows when to reach for it.

Send Email
Sends an email mid-call from a mailbox you connected under Mailboxes. Off by default. Pick the mailbox on the gear; the agent confirms the address before sending.
Schedule Callback
When the agent cannot reach a person (no answer, busy), it logs a callback for the team - it does not dial anyone. Cap the pending callbacks per caller (1 to 3) so repeats do not read as spam; past the cap, an alert email goes out instead. This cap counts callback tasks per caller; it is separate from a campaign’s retries (Call limits).
Detect Live Sentiment / Detect Live Frustration
Not active on calls yet. The setting saves, but its prompt does not reach the agent during a call, so turning it on changes nothing today. Leave both off; to de-escalate an upset caller, write the rule in the personality prompt.
Add Participant
The agent dials an extension or number the caller gives and adds that person to the live call; nobody is dropped. Leave dialing outside the PBX off unless you need it.
Make Call Private
Hides this call from the rest of the team when the person the agent assists asks to keep it private.
Contact verification
Confirms who is on the line before you share account details: a one-time code by SMS or email, or - for a contact you already hold - the number you called them on plus a field they must know (for example ΑΦΜ). On widget, Messenger and Viber the caller stays anonymous until it succeeds.

Callers often say “one moment” or «περιμένετε ένα λεπτό» and go to find a document. Without this tool the line goes quiet, Handle Silence counts it as silence, and it can nudge or end the call while the caller is still looking.

With Wait for Caller on, the agent answers briefly (“Of course, I’ll wait.”) and stays silent. Handle Silence does not check on the caller until they speak again or the limit passes. After that, Handle Silence works as usual: it nudges, then forwards or ends the call as you configured it.

Where
Skills - Tools & Workflows - Built-in Tools - Wait for Caller. Turn on the toggle; the gear opens Wait for Caller - Tool Settings.
Maximum wait (seconds)
The longest the agent waits in silence after the caller asks it to wait: 1 to 300 seconds, 120 by default. If the caller names a shorter time (“half a minute”), the agent waits that long.
Prompt
When to use the tool. The default covers «περιμένετε ένα λεπτό» and “hold on”; add your own phrases if your callers say it differently.

It is off by default, applies to phone calls, and works together with Handle Silence: with Handle Silence off there is nothing to hold, so the tool changes nothing.

When a caller talks over the agent, the agent stops mid-sentence. Without this setting the agent still remembers the whole sentence it had started, as if the caller heard all of it, and it is not told that it was cut off. With Interruption by the caller turned on (Skills - Tools & Workflows, below the built-in tools), the agent remembers only the words the caller actually heard. Before its next reply it gets one short note with what it was going to say, what the caller heard and what the caller said. The note is used for that one reply only.

It is off by default and applies to phone calls only. The call transcript shows a step like ⏸ Διακοπή: ακούστηκε έως «…» where the caller cut in.

To write your own note, fill in the text box under the switch; leave it empty to use the built-in text. You can use three placeholders:

{full_text}
What the agent was going to say.
{spoken_text}
The part the caller heard.
{user_text}
What the caller said when they interrupted.

For example: You were cut off. The caller heard only “{spoken_text}” and said “{user_text}”. Answer them first, and repeat the rest only if it still matters.

Sometimes the agent needs a few seconds before it starts to answer (a long question, a busy AI model). With Soft timeout turned on (Skills - Tools & Workflows, below Interruption by the caller), the agent says one short line such as “One moment, let me check.” when its reply has not started a few seconds after the caller stopped talking. When the reply comes quickly, it says nothing.

It is off by default and applies to phone calls only.

Seconds before the line
2 to 10 seconds of silence after the caller stopped talking. 3 suits most agents.
Wait lines
Optional, one per row, up to 10 lines of 80 characters. The agent uses them in turn and never says the same line twice in a row. Leave it empty to use the built-in lines in the agent’s language.

The same lines are also used when a slow tool, workflow or knowledge lookup is still running after 6 seconds, so the agent sounds the same in every wait. It never speaks over the caller, and the line is cancelled as soon as the reply starts. The call log shows a step 🕐 Soft timeout when a line was said.

How it differs from the other two waiting aids:

  • the message before a tool is said every time that tool starts, fast or slow;
  • the keyboard sound (wait music) plays under any pause, with no words.

The gear prompt is the contract: when to send the SMS, when to stay silent. Vague prompts fire at the wrong moment.

The personality prompt should mention the tool in one line. Otherwise the model may never choose it, even when the toggle is on.

Add MCP if you have a remote tool server. Choose transport (for example Streamable HTTP) and auth (None, Bearer token, or OAuth). OAuth redirects to connect, then you return and select tools. Only the tools you enable are sent to the AI. An MCP that is added but not connected will not run.

Company-wide Claude / ChatGPT connections (Edit with AI) are Integrations - AI Connections, not this list.

Should I enable every built-in tool for my agent?

No. Enable only what this agent’s job requires, because every extra tool is another way for it to go off-script. The Enabled list under Skills - Tools & Workflows should stay a short list you can explain out loud, not dozens from the catalog.

Why does the Required checklist mention Handle Silence on my agent?

Handle Silence is set to forward the call, but Transfer is not ready. Either configure Transfer under Call handling, or change the silence action to End the call.

Skip Turn is on, so why does the agent still talk over me when I read a phone number?

The Skip Turn settings (the gear) have no prompt, or the prompt is vague. Open the gear, use the Phone number template (or Order number / Product code), set the timeout and max skips, and Save. Then mention Skip Turn in one line of the personality prompt so the model knows when to use it.

Why does my workflow run in the editor but never during a call?

Either the workflow is not attached and toggled on for this agent under Skills - Tools & Workflows, or the prompt never tells the model when to use it. Attach it, turn its toggle on, and add one line to Instructions saying when to run it.

Why does the location in the SMS form show the right area but the wrong street?

The position comes from the caller’s phone: the form asks for GPS first, and if GPS does not lock within 15 seconds, common indoors, it uses Wi-Fi or mobile network positioning, which can be off by hundreds of meters. The address under the map is the street nearest those coordinates, from OpenStreetMap, so the area is right and the street is not. When the position is approximate, the form says so (for example “±300 m”) and asks the caller to drag the pin to the right spot, which corrects both the coordinates and the address. The form submits the coordinates; the address is a guide.