Reach out when buyers show intent

Demand-gen / GTM engineer / GTM ops / founder-led sales · Assemble in ~15 min; first run in <2 min once installed.

GTMEdit flowon X◐ readiness 43

step 1ready

Define which signals predict buying (ICP signal model)

niche-signal-discovery12 alts

step 2connector

Detect real-time intent across LinkedIn + open web

⚡ LinkedIn Sales Navigator

step 3partial

Operate the 6sense intent + scoring layer

sales-6sense7 alts

step 4partial

Triage in-market accounts into a worklist/tracker

outreach14 alts

step 5ready

Build the signal-triggered sequence/cadence

outreach-sequence-builder3 alts

step 6ready

Personalize per-prospect copy tied to the signal

deal-outreach8 alts

step 7connector

Send, track, and follow up

⚡ bring your own tool

Did this recipe work for you?

What you get

▸Intent watchlist (who is in-market right now, why)
▸Triggered outreach drafts tied to each signal

tools: signal source needed (LinkedIn/news monitoring; intent data tools are $$$)

tokens: ~$2-4 per sweep

≈ depends heavily on signal source

What stays yours

· Connect/choose the signal source
· Approve sends

The problem

When prospects show real-time buying signals (funding, job posts, tech changes, LinkedIn activity, site visits, post-engagers), I want an agent to detect intent across LinkedIn and the open web and only fire outreach when someone is in-market, so I stop spray-and-pray and contact people exactly when they're likely to buy.

Readiness

The downstream half assembles cleanly from the library — signal definition (#256), cadence design (#75), and personalized copy (#441) are all ready. But the defining capability of this job, live intent detection across LinkedIn and the open web, is a byo-tool hole: every real signal feed (6sense/Bombora, Clay/Apollo, Sales Nav, funding/job-post news, site-visitor de-anon) plus the sender and the state/scheduler must be brought. The skills tell you what to watch for and what to say — they don't see the signals or push the send.

Cybernetic layers

Memorybyo

Must remember which accounts already fired, which signals were seen and when, and suppression windows so the same intent event doesn't re-trigger. No candidate persists this — bring a state store (the Notion tracker from #602, or a CRM/Postgres) keyed by account+signal+timestamp.

Background / schedulingbyo

Intent detection is inherently unattended and continuous — the agent must poll signal feeds on a schedule (e.g. hourly/daily) and fire only on fresh signals. Needs a cron/scheduler (Claude Code schedule routine or external cron) wired to the intent APIs; no candidate provides the loop.

Harnessready

Claude Code runs the skills as a pipeline — invoke signal-definition (#256) once to set rules, then run detect→triage→sequence→personalize on each scheduled pass. Harness is ready; the data connectors and scheduler are the BYO gaps.

Click any step to see what fills it — and the ranked alternatives you can swap in.