> ## Documentation Index
> Fetch the complete documentation index at: https://help.pipelime.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Agent alerts and what to do about them

> Every banner an agent can raise, what caused it and the one action that clears it.

export const UiPath = ({children}) => {
  const parts = String(children).split(/\s*(?:→|>)\s*/).filter(Boolean);
  return <span className="pl-path">
      {parts.map((part, index) => <span key={`${index}-${part}`}>
          {index > 0 && <span className="pl-path__sep">→</span>}
          {part}
        </span>)}
    </span>;
};

export const Screenshot = ({id, alt, caption, frame = 'browser', url, marks = [], lang = 'en', workspace = 'northwind-outbound'}) => {
  const src = `/images/screenshots/${lang}/${id}.png`;
  const chrome = url ? `app.pipelime.ai/${workspace}${url}` : 'app.pipelime.ai';
  return <figure className={`pl-shot pl-shot--${frame} not-prose`}>
      <div className="pl-shot__frame">
        {frame !== 'bare' && <div className="pl-shot__bar">
            <span className="pl-shot__dots">
              <span className="pl-shot__dot" />
              <span className="pl-shot__dot" />
              <span className="pl-shot__dot" />
            </span>
            <span className="pl-shot__url">{chrome}</span>
          </div>}
        <div className="pl-shot__media">
          <img src={src} alt={alt} loading="lazy" />
          {marks.map(mark => <span key={mark.n} className="pl-shot__mark" style={{
    left: `${mark.x}%`,
    top: `${mark.y}%`
  }} aria-hidden="true">
              {mark.n}
            </span>)}
        </div>
      </div>
      {caption && <figcaption className="pl-shot__caption">{caption}</figcaption>}
    </figure>;
};

Every problem an AI agent can flag turns into a banner across the top of its **Dashboard** tab. This page covers all of them: what triggers each one, and the one action that clears it.

<div className="pl-availability">
  <div className="pl-availability__row">
    <div className="pl-availability__label">Where</div>
    <div className="pl-availability__value"><span className="pl-path">Outreach<span className="pl-path__sep">→</span>AI Agents</span></div>
  </div>

  <div className="pl-availability__row">
    <div className="pl-availability__label">Your role needs</div>
    <div className="pl-availability__value">Read access to AI Agents <code>read-campaigns</code>. Admin, Member and Viewer have it by default.</div>
  </div>

  <div className="pl-availability__row">
    <div className="pl-availability__label">To create or change</div>
    <div className="pl-availability__value"><code>create-campaigns</code> to add one, <code>update-campaigns</code> to change one, on top of the permission above.</div>
  </div>

  <div className="pl-availability__note">If you cannot find this in your sidebar, your workspace may have a custom menu configuration. Contact support and we will check it for you.</div>
</div>

## Where they show up

Open the agent from <UiPath>Outreach → AI Agents</UiPath> and stay on its **Dashboard** tab. The banners sit at the top of it, and only one thing can ever sit above them: the first-leads card, on an agent you launched yourself. Each banner carries an icon, a title, a line explaining it and the button that deals with it.

Two other tabs repeat part of the strip, as the same cards. **Settings** raises the missing sending channel and the goal with no call-to-action, worked out again from the form in front of you. **Sources** raises the two sourcing pauses: out of credits, and a source stopped at its company limit. Both tabs also carry a colored dot while they have something to show. Everything else appears on the **Dashboard** tab and nowhere else.

The banners arrive in a fixed order that has nothing to do with severity: the sending channel and WhatsApp reach first, then email accounts and tags, then outreach and the goal, then sourcing's own health, and the no-matching-leads warning last. Color is what tracks severity. Three are tinted as errors: no sending channel, the missing goal, and sourcing out of credits. Six are warnings. Two sit in the mildest tint, kept for information rather than a problem: no tag linked, and every source having finished.

When an agent has nothing to flag, the strip isn't there at all: no empty box, nothing. And a problem the product learns to raise later, which this page would not cover, still gets a card of its own headed **This agent needs attention** rather than failing silently.

<Screenshot id="ai-agents/agent-alerts-and-issues--stack" url="/ai-agents/12" alt="Two stacked banners at the top of a fresh agent's Dashboard tab: 'No sending channel connected' with a Connect senders button, and 'Outreach is paused while sourcing runs' with a Resume outreach button" caption="A workspace on its first day: nothing to send from (1), and outreach off while the source runs (2). A first agent whose goal has no destination yet raises a third banner in the same strip." marks={[{ n: 1, x: 88, y: 26 }, { n: 2, x: 88, y: 72 }]} />

## The two the product won't let you ignore

Two banners are marked blocking by the checklist behind them: **No sending channel connected** and **Goal is missing its call-to-action**. Those two, and only those two, follow you to the agent's **Settings** tab as the same alert cards, and put the dot on that tab.

Red is not the same thing as blocking. **Sourcing paused — out of credits** is tinted the same error-red without being one of the two: it never appears on **Settings**, and goes to the **Sources** tab instead.

### No sending channel connected

The agent has nothing to send from: no email account, no LinkedIn account, no WhatsApp account. **Connect senders** (1) opens the **Add senders** overlay on top of the page you're on. An agent set to rotate across every email account in the workspace still shows this while the workspace has none. Rotating across an empty list sends nothing.

### Goal is missing its call-to-action

The agent knows what it's trying to achieve but has nowhere to send anyone. **Set up goal** opens the **Goal** section of **Settings**. Fill in whichever field the chosen goal needs (a meeting link, a phone number, a website address or your own custom description), and the banner clears.

Whichever of the two applies also opens its own section the moment you land on **Settings** (**Channels** for the missing channel, **Goal** for the missing call-to-action), so you're already looking at the field to fix.

<Screenshot id="ai-agents/agent-alerts-and-issues--settings-banners" url="/ai-agents/12?tab=settings" alt="The agent's Settings tab with the 'No sending channel connected' alert card at the top, above the Channels section already expanded beneath it" caption="Settings raises only these two. The agent here already has its goal filled in, so only the channel banner is showing." />

## Email accounts, tags and WhatsApp reach

### Some senders need attention

One or more of the email accounts this agent can send from is in an error state. **Review senders** (1) leaves the agent and opens **Email Accounts**, where the errored row is waiting. If the agent rotates across every account in the workspace, the whole roster is checked; if it's set to a specific list, only those accounts count.

<Screenshot id="ai-agents/agent-alerts-and-issues--sender-attention" url="/ai-agents/45" alt="An agent dashboard showing the 'Some senders need attention' banner with a Review senders button" caption="Representative example: one mailbox this agent sends from has stopped working." marks={[{ n: 1, x: 88, y: 50 }]} />

### No tag linked to this agent

Without a tag the agent can't track its own leads, funnel or metrics.

<Warning>
  **Link a tag** does switch you to the **Settings** tab, but the section name it passes doesn't match any of the five the tab actually recognizes, so nothing expands or scrolls once you land there. Open **Advanced** yourself: **Audience tags** is the block to look for.
</Warning>

### No WhatsApp-reachable leads yet

This one is judged only on the agent's own page. It needs a live count of how many of the agent's leads have a phone number, and only that page fetches it. It fires only for an agent that sends on WhatsApp alone, once it actually has leads and none of them has a phone number. **Review channels** opens the **Channels** section of **Settings**: adding an email or LinkedIn channel, or sourcing leads with phone numbers, both clear it.

### WhatsApp accounts can't send right now

Every WhatsApp account on the agent is out of budget for today: a daily limit, warm-up, or its sending window. **Review accounts** opens the **WhatsApp Accounts** page. There's nothing to fix here: sending resumes on its own once the budget refreshes.

<Note>
  Neither WhatsApp banner can reach an agent with no WhatsApp number attached to it, and the agent only offers WhatsApp as a channel once your plan includes WhatsApp numbers, or the agent already has one attached from before. On every other workspace these two never appear.
</Note>

## Sourcing and outreach

### Outreach is paused while sourcing runs

Leads are arriving and nobody is contacting them. It only fires while at least one lead source is genuinely producing: actively running, or resting for the rest of the day after hitting its own daily cap. A source that's merely queued and hasn't started yet doesn't count, so a workspace whose first source is still waiting its turn shows no banner here even with outreach off.

**Resume outreach** acts in place: no page change, no dialog. It shows a spinner while it works, then confirms with **Outreach resumed**. If it fails, a general error message appears and the banner stays. It resumes outreach only. Nothing about the lead sources changes.

### Sourcing paused: out of credits

At least one of the agent's lead sources stopped because the workspace ran out of credits. **Get more credits** opens the subscription dialog over the page, so you keep your place. What happens to a paused source once the workspace has credits again is covered on [your first agent: what to do when something looks wrong](/en/getting-started/troubleshooting-your-first-agent).

### Sourcing reached its company limit

A source hit the company limit you set for it. The exact same banner and button also appear when a source just finished: it did everything it could and stopped on its own, rather than being capped. The [**Setup checklist**](/en/getting-started/setup-checklist) in the side rail tells the two apart without you leaving the tab: its **Sourcing running** row counts as done once every source has finished, and asks for attention while one is stopped at its limit. The **Sources** tab repeats this banner for the capped case only, not for sources that finished.

**Increase limit** opens a small editor for that source's cap, right next to the button, so you never leave the tab. If more than one of the agent's sources is capped or finished, the button can't know which you meant, and it switches you to the **Sources** tab instead.

### No leads matched yet

Sourcing analyzed companies and none of them cleared your criteria. Like the WhatsApp-reach warning above, this one is only ever judged on the agent's own page. **Adjust targeting** opens the **Sources** tab.

## The issue count next to an agent

On the **AI Agents** list, each agent card carries a small pill counting its open issues, next to its status. It turns error-red if any of them is error-level and amber otherwise: an agent whose only issue is an informational one still gets the amber pill, never a third color. Hover it and a list appears, one line per issue with a colored dot.

That count is worked out from what the list already loaded, and the list loads a slimmed-down agent: no channel accounts, no sending capacity, no live lead counts. Five of the banners above can never reach it: **No sending channel connected**, **Some senders need attention**, both WhatsApp ones, and **No leads matched yet**. What it can count is the missing tag, the missing goal, outreach paused while sourcing runs, and the three sourcing states. So the number can go up the moment you open the agent, and the **Dashboard** tab is the one to trust.

<Note>
  The hover list is keyed by an older set of names than the checklist emits today, so half of the lines it can show read as raw text with the dashes turned into spaces: "Missing goal cta", "No tags", "Outreach paused while sourcing". The three sourcing lines are the ones that read properly, and they are also the only place the product separates a source stopped at its limit ("Lead sourcing reached its company limit") from sources that have all finished ("All lead searches have finished"), which the banner wraps into one wording. Trust the banner titles on this page over what the hover list says.
</Note>

## Related

<CardGroup cols={2}>
  <Card title="No sending channel connected" icon="send" href="/en/troubleshooting/agent-not-sending">
    Every reason an agent sends nothing, in the order worth checking.
  </Card>

  <Card title="No leads matched yet" icon="search" href="/en/troubleshooting/no-leads-found">
    What to change when sourcing runs and nothing matches.
  </Card>

  <Card title="Fix a sender in Error: reconnect" icon="mail-warning" href="/en/email-accounts/reconnect-a-sender">
    Every message an errored mailbox can show, and what Reconnect actually does.
  </Card>

  <Card title="What consumes credits" icon="coins" href="/en/billing/what-consumes-credits">
    What actually costs a credit, for when sourcing pauses to ask for more.
  </Card>
</CardGroup>