Skip to content
Agent Library

Agents that do real sales work

Pick an agent, give it the task, and review the result before anything is saved. Every run is logged.

See pricing
Agentic SystemsAgent LibrarySearch…⌘KYour workspacePR

Agent Library

Pre-built agents that do real sales work — pick one, give it the task, review before anything is saved.

Find
Lead Scraping Agent

Give it an ICP → it discovers matching people on the web, deduped against your CRM.

Uses the webRun
Understand
Research Agent

Live web research → a structured brief with sources, ready to save to the record.

Uses the webRun
Engage
Outreach Agent

Turns a goal + CRM context into a personalized email, ready to send + log.

CRM data onlyRun
Advance
Pipeline Guardian

Scans your open pipeline for stalls and proposes the next best action per deal.

CRM data onlyRun
Close
Meeting Prep Agent

One-page brief from everything in the CRM — contacts, deals, touchpoints, quotes.

CRM data onlyRun
Build your own
Agent Studio

Drag blocks: a trigger, the records it needs, one AI step, an action.

Dry run firstOpen
Example
The agents

One agent for each stage of the deal

Each agent owns one stage of the deal and shows the data it uses.

  • FIND

    Lead Scraping Agent

    Give it an ICP. It finds matching people on the web and skips anyone already in your CRM.

    Uses the web
  • UNDERSTAND

    Research Agent

    Runs live web research and returns a structured brief with sources, ready to save to the record.

    Uses the web
  • ENGAGE

    Outreach Agent

    Turns a goal and your CRM context into a personalized email, ready to send and log.

    CRM data only
  • ADVANCE

    Pipeline Guardian

    Scans open deals for stalls and suggests the next best action for each one.

    CRM data only
  • CLOSE

    Meeting Prep Agent

    Builds a one-page brief from everything in the CRM: contacts, deals, touchpoints and quotes.

    CRM data only
  • BUILD YOUR OWN

    Agent Studio

    Combine a trigger, the records it needs, one AI step and an action.

    Dry run first
How a run works

You stay in charge of every run

  1. Pick an agent.
  2. Give it the task.
  3. Review the result.
  4. Save it. The run is logged.
Built to be trusted

Clear about what each agent does

  • Review first

    Nothing is saved until you approve it.

  • Clear data sources

    Every agent shows whether it uses the web or only your CRM data.

  • Every run logged

    See what each agent did, and when.

Agent Studio

Build an agent for your own process

Drag four blocks into place, then test with a dry run before it runs for real.

  1. Trigger Choose what starts the agent.
  2. Records Pick the records it needs.
  3. AI step Add one AI step.
  4. Action Choose what happens with the result.
Questions

Good to know

No. You review each result before it's saved.

Give an agent its first task

20 minutes. We'll run each agent live and show you the review step.

See pricing