Built-in agents
Built-in agents are agents FunnelStory builds, tunes, and runs for one specific job. You don't design their steps; at most you turn on a setting. Each reads real evidence from your workspace, reasons through it, and ends with an action or a draft for a person to review. See Built-in and Outcome Agents for how they differ from the agents you build.
Needle Mover investigation
Investigates an open needle mover — the conversations, relationship history, commercial terms, and prediction behind it — and drafts an Action Card with the recommended next step and, when it helps, a ready-to-send email. It runs two ways:
| Mode | How it runs | Where |
|---|---|---|
| Assist | On demand: ask it to look into the needle mover, and it investigates and drafts the card in the conversation | The Investigation and Triage chat on a needle mover's detail view — see AI summaries |
| Auto | On a schedule: it investigates open needle movers your team hasn't acted on and drafts a card before anyone asks | Autosuggested action cards, turned on under Configuration → Features → Needle Movers — see Autosuggested action cards |
Both modes use the same investigation and land in the same place: an Action Card assigned to the account's CSM.
Meeting prep
Before each meeting with a customer account, an agent builds a briefing — the account's health, recent activity, feature requests, notes from earlier meetings, and open action items — and sends it to you in Slack about 30 minutes before the meeting starts. It runs automatically once your calendar and Slack are connected; there's nothing to turn on. See Meeting prep for setup.
Related
- Agents — the three kinds of agents
- Outcome Agents — agents that work toward a retention or expansion goal
- Action Cards — where built-in agents' drafts land