A free Claude Code skill that turns your RB2B-identified website visitors into a routed, tiered outbound playbook before the 48-hour intent window closes.
RB2B Outbound Triggers is a Claude Code skill: a markdown playbook that turns a list of RB2B-identified website visitors into a complete, tiered outbound plan. Give it your visitor data (name, title, company, LinkedIn, pages visited, session length) and it sorts every visitor into one of four tiers, builds the enrichment waterfall to find a verified email, designs the exact touch-by-touch sequence for each tier, writes the Slack alert your SDR should get, and lays out the Clay automation that connects RB2B to your sequencer.
The video's claim is that it "takes identified website visitors and turns them into an actual outbound sequence, routed and ready, inside the 48-hour window before their interest goes cold." That holds up. Tested against 5 invented sample visitors for a fictional B2B SaaS, it correctly split them into Tier 1 Hot (2 decision-makers on buying-stage pages), Tier 2 Warm (1 influencer on mid-funnel pages), Tier 3 Monitor (1 visitor from a company too small to be in the ICP), and Tier 4 Exclude (1 visitor flagged as a named competitor and suppressed before any outreach). Every tier got the right treatment: no phone touches wasted on Tier 2, no enrichment run on Tier 3 or 4, sequence copy that references the visit contextually instead of sounding like surveillance.
One honest caveat: this is a playbook generator, not a live system. It does not connect to your real RB2B account, Clay workspace, Slack, or sequencer and push anything automatically. It hands you the routing rules, the sequence content, the Slack payload format, and the Clay workflow spec, ready to wire into the tools you already run. "Routed and ready" describes the plan it produces, not something that fires on its own the moment you install it. The 48-hour, "50% intent decay" framing the skill uses is also its own stated assumption, not a cited external statistic, worth knowing if you repeat that number on camera.
Anyone running RB2B (or about to) who has visitors getting identified but no consistent process for what happens next. This replaces the "someone forwards a Slack alert and maybe an SDR follows up eventually" routine with an actual tiering system, so a VP of Engineering who just spent 6 minutes on your pricing page gets a 15-minute-SLA sequence instead of getting lost in the same queue as someone who read one blog post.
This copies the whole gtm-skills library onto your machine. You only need one folder out of it.
git clone https://github.com/LeadMagic/gtm-skills.gitThis drops just the rb2b-outbound-triggers skill into the folder Claude Code checks for project skills.
mkdir -p .claude/skills
cp -r gtm-skills/skills/outbound/rb2b-outbound-triggers .claude/skills/rb2b-outbound-triggersNote: the repo also documents a one-line ./install.sh and a gh skill install command. Both rely on a gh skill extension that was not present on the test machine, and the one-line installer reported success paths but wrote zero files instead of falling back. The plain copy above is the path that actually worked, and it is the exact fallback method the repo's own installer uses internally.
Sample input used in testing, 5 invented RB2B-identified visitors for a fictional DevOps SaaS:
Priya Nandakumar, VP of Engineering, Meridian Freight Corp (600 employees)
visited /pricing, /docs/api-auth, /case-studies, 6m42s, repeat visit
Dana Okafor, Head of Platform Engineering, Solera Freight Systems (900 employees)
visited /pricing, /demo, 4m58s
Tomas Alvarez, DevOps Lead, BrightPath Logistics (310 employees)
visited /product/incident-response, /case-studies, 3m10s
Wei Chen, Site Reliability Engineer, Forge Analytics (45 employees)
visited /blog/observability-trends, 1m05s
Sam Wexler, Sales Engineer, MonitorHive (a named competitor)
visited /pricing, 0m48sAbbreviated real output, following the skill's own routing matrix:
Priya Nandakumar -> Tier 1 Hot (ICP 92/100) -> 7-touch sequence, Slack alert, 15-min SLA
Dana Okafor -> Tier 1 Hot (ICP 95/100) -> 7-touch sequence, Slack alert, 15-min SLA
Tomas Alvarez -> Tier 2 Warm (ICP 68/100) -> 5-touch nurture, no phone touch
Wei Chen -> Tier 3 Monitor (ICP 35) -> logged in CRM, no outbound
Sam Wexler -> Tier 4 Exclude (competitor) -> suppressed, added to blacklistThe full deliverable also included a filled Slack alert template, a step-by-step Clay workflow spec with error handling, a measurement dashboard with this batch's actual numbers, and a privacy guardrails section. Ran the skill's own validator against the finished deliverable:
$ python3 scripts/check-output.py deliverable.md
PASS, RB2B Outbound Triggers deliverable meets minimum checks./install.sh or gh skill install fails with unknown command "skill" for "gh". Your GitHub CLI does not have the gh skill extension the repo assumes, and the installer will not fall back to copying the files on its own. Skip it and use the plain cp -r copy in Step 2 instead, it lands in the same place.
scripts/cc-gtm.py (the "curated installer") crashes with a SyntaxError about an f-string and a backslash. That script needs Python 3.12 or newer. If you're on an older Python 3 (macOS ships 3.9 by default), skip it and use the cp -r copy method instead.
You want the alerts and sequences to actually fire, not just get written. They won't on their own. Wire the Slack alert format into your real RB2B-to-Slack integration, and follow the skill's own Clay workflow spec to connect RB2B's webhook to your enrichment waterfall and sequencer.
A visitor's company doesn't clearly fit your ICP. That's what Tier 3 is for. The skill deliberately withholds enrichment and outbound from anything under your ICP floor, don't force a sequence onto a visitor that scored low, log it and move on.
Built by LeadMagic (https://github.com/LeadMagic/gtm-skills), MIT license, 53 stars at test time. The video's description holds up: it produces a real, routed outbound plan from identified visitors. One clarification not in the video: it is a playbook generator you wire into RB2B, Clay, Slack, and your sequencer yourself, it does not connect to those tools and act on its own.
This is one agent. The 7-Day AI Native Challenge is how you learn to build the next ten yourself, live, with Dany.
Join the Challenge →Too busy to build it yourself? Focentra AI will build, install and train your team on it. Priced on the call.
Talk to Focentra AI