Lead Briefing
New-lead summariesThis agent reads each new lead and writes a short summary for the rep: who they are, what they asked for, and what to do first.
Worxley Agents
Give the busywork to an agent.
Agents read your leads, deals and customers when something happens, then summarise it for you, flag risk or prepare the next step. They stay inside the limits you set, and they ask before they act.
New leadPriya Nair
Summary readyFor the owner
The team
A team of agents, each with one role
Pick one, switch it on, and it starts on the next matching event. You can change its instructions before you publish.
Deal Risk Monitor
Deal monitoringThis agent watches open deals for silence, slipping dates and stalled stages, and flags the ones that need a person today.
Customer Health
Account monitoringThis agent reads a customer's tickets, notes and recent activity and tells you whether the account is healthy or drifting.
Stale Lead Watch
Alerts for untouched leadsThis agent finds leads nobody has touched in a while and tells the owner before they lose interest.
Follow-up Planner
Follow-up planningThis agent prepares the next follow-up for a lead or deal and asks you before it books or sends anything.
Your own agent
Drafting helpDescribe the task in plain words. Worxley drafts the agent for you. Nothing runs until you publish it.
How it works
It starts on an event, then follows its instructions
There is no hidden step. Every run follows the same four moves, and you can read what it did afterwards.
Lead Briefing
New-lead summaries- Event received: new leadPriya Nair
- Woke up Lead BriefingJust now
Guardrails
Useful, and kept within limits
An agent only does what its tools, its budget and its approvals allow.
Tool bundles
Each agent gets a fixed set of tools. A read-only bundle cannot change anything, whatever the instructions say.
Approval before action
Logging an answer, requesting a callback or booking a meeting can each be set to wait for a person.
Its own budget
Every agent has its own usage limit, so one runaway agent cannot use up the rest.
Draft first, publish second
New agents start as drafts. Someone with publish permission reviews the goal, instructions and access, then activates it.
Triggers
Pick the moment, the agent does the rest
An agent can be started by any of these events.
- A new lead is createdLead
- A lead's status changesLead
- A new customer is createdCustomer
- A customer moves to a new stageCustomer
- A customer raises a support ticketCustomer
Your agent- A summary of each new lead for the owner
- An alert when a deal is at risk
- A check on how a customer is doing
- The next follow-up, prepared
- A drafted reply, ready for approval
Questions teams like yours ask
An agent is a small AI worker with one role. It starts when something happens in your CRM, reads the record, and then reports back or asks to take the next step. It works inside the tools, budget and approvals you give it.
Only if you let it. Read-only agents cannot change anything. Agents that can log answers, request callbacks or book meetings can each be set to wait for your approval first.
Only what its tool bundle allows, such as the lead or customer, its deal and its notes. A support agent reads the customer's own records and nothing more.
A workflow follows the exact steps you drew. An agent is given a goal and instructions and works out what to look at. Many teams use both: workflows for fixed steps, agents for reading and judging.
Yes. Start from one of the agents above, or describe the task in plain words and Worxley drafts one for you. It stays a draft until someone with publish permission activates it.
Anyone with publish permission. Agents drafted by the AI Copilot or an AI client are marked as drafts and wait for a person to review them.
APPLY FOR COHORT 1
You've met the agents. Put one to work.
Switch on one of our agents, or describe your own, and see what it flags on your next lead.