Use Case

Lead generation that runs as an operating system.

An AgentLed MCP workspace keeps ICP context, lead research, approvals, outreach, and conversion analytics in one loop — so every agent run learns which prospects are actually worth pursuing.

agentled.app / lead sourcing run
Example run report

Lead Sourcing - 1,240 scanned · 42 outreach emails sent

Agents searched six source channels, normalized company evidence, scored 186 ICP-fit leads, and sent 42 approved outreach emails with source-backed messaging.

Best source

LinkedIn

Profiles Scanned

1,240

ICP-Fit Scored

186

Sources Used

6

Outreach Sent

42

Source coverage

found / qualified

LinkedIn

best seniority match

72 / 18

Twitter/X

buying-trigger signals

34 / 9

Reddit

pain-point threads

21 / 5

Product Hunt

recent launches

18 / 6

Crunchbase

funding and growth

31 / 8

Web Scraping

company-site evidence

56 / 12

Top segment

B2B SaaS operators showing active AI automation intent and recent process-change signals.

Outreach status

Approved emails sent42
HubSpot records updated42
Replies pending42

Northstar RevOps

VP Revenue Ops

Score: 91email sent

LedgerBase

Head of Finance Automation

Score: 88email sent

MosaicCare

COO

Score: 84approval queued
01The Problem

ICP context lives in heads

Target accounts, disqualifiers, buying triggers, and objection notes sit across docs, calls, CRM fields, and founder memory.

Lead research goes stale

Static lists miss role changes, new funding, product launches, hiring signals, and the reasons a lead became relevant now.

Outreach is disconnected from outcomes

Replies, meetings, rejections, and sales notes rarely update the research system, so every new campaign repeats old guesses.

02How It Works
01

Load ICP and market context

AgentLed MCP turns your segment rules, customer notes, exclusion criteria, and buying triggers into reusable prospecting context.

02

Maintain account and lead research

Agents monitor LinkedIn, Apollo, Hunter, company sites, news, and CRM history to keep lead records current instead of one-off.

03

Score fit with evidence

OpenClaw compares each lead against ICP, intent signals, timing, and past outcomes, then records the reasoning behind the score.

04

Draft outreach for approval

Codex prepares message variants with cited research, suggested angle, and next-best action so humans can approve, edit, or reject.

05

Run outreach and handoffs

Hermes sends approved sequences, routes replies, books meetings, and passes qualified opportunities into the operating workspace.

06

Close the analytics loop

Replies, meetings, conversion quality, objections, and rep feedback update the Knowledge Graph so the next run targets sharper leads.

03Results

43

approved ICP-fit leads per run

156

research-backed profiles maintained

1

workspace for research, approvals, outreach, and analytics

04Connected Tools
AgentLed MCPCodexOpenClawHermesLinkedInTwitter/XRedditProduct HuntCrunchbaseWeb ScraperHunter.ioApolloGmailSlackHubSpotKnowledge Graph

Turn this operation into an agent-run workspace.