agentleFS
NadavShanun-designNadavShanun-design/opengtm/skills/multichannel-sequences/SKILL.md

multichannel-sequences

Skill4 starsChanged 29 days agoMITView on GitHub

Use when designing or auditing an outbound cadence - how many touches, how far apart, which channels (email, LinkedIn, phone, SMS), what each touch says, branching on behavior, and when to stop. Covers sequence architecture, channel roles, spacing, reply/branch logic, and exit rules. For the email copy itself see cold-email-writing; for LinkedIn mechanics see linkedin-outreach.

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 3 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    SMS: - only with prior relationship/opt-in.

    line 55
  2. Distinctive:

    Do stop on reply and honor opt-outs instantly.

    line 78
  3. Distinctive:

    Don't: run 15 touches - diminishing returns become brand damage.

    line 80

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Multichannel sequences

Overview

A single cold touch is a coin flip; a well-built sequence is a system. The job of a cadence is to earn attention across several relevant touches without becoming noise, and to stop the instant it should (reply, meeting, or disinterest). Sequence = architecture; copy (cold-email-writing) and channel mechanics (linkedin-outreach) hang off it.

Design the architecture, not just the emails

A touch has four decisions: channel, timing, angle, and ask. Vary the angle across touches (new value each time - a different pain, a proof point, a resource) rather than "just following up." Every touch must stand on its own; a prospect may only ever see touch #3.

A solid default cadence (tune to your motion)

#DayChannelAngle
10EmailTrigger/relevance + one ask
21-2LinkedInView profile; connect (no pitch) or comment on a post
33-4Email (reply to #1)New angle + proof point
46-8LinkedIn DM / callShort, human, reference prior touch
510-12EmailDifferent pain or a resource (no ask fatigue)
614-18EmailPolite break-up ("should I close this out?")

That's a blended ~6-8 touch starter. Scale the total by deal size (across all channels): roughly 8-10 for SMB over ~2 weeks, 10-14 for mid-market, and 12-16 for enterprise over 3-4 weeks. The follow-ups do the heavy lifting - one large study found steps 2-6 drove ~59% of all replies (step 3 alone a third of email-sourced meetings), so the first touch is rarely where the reply comes from. Diminishing returns are real, but they come from email fatigue, not touches per se: cap each email wave at ~3 and switch channels rather than send a 4th near-identical email (4+ emails in a row sharply raises spam/unsubscribe rates).

Channel roles

  • Email - the workhorse: carries the detail and the ask.
  • LinkedIn - warms and humanizes: profile view, genuine comment, soft connect. (Respect LinkedIn's invite limits - see linkedin-outreach.)
  • Phone/voicemail - highest intent, best after a prospect has seen your name.
  • SMS - only with prior relationship/opt-in; powerful but easy to overuse.

Sequence the channels so each one references the last - a connected story, not four strangers shouting the same thing.

Branch and exit logic

  • Reply (any) → pull them out of automation immediately; a human (or reply-handling-and-objections) takes over. Never let a sequence keep firing at someone who answered.
  • Opened repeatedly, no reply → a lighter "is this a priority?" nudge.
  • Hard no / unsubscribe → stop, suppress, honor it instantly.
  • No engagement by last touch → exit to a long-term nurture, not the trash.

Pace for deliverability

Spread sends evenly across the day and across inboxes; never burst a whole list at once. Warmup and daily caps still apply (see email-deliverability). An even, paced send protects the domain that the whole sequence depends on.

Do / Don't

  • Do give every touch a fresh angle and make each one self-contained.
  • Do stop on reply and honor opt-outs instantly.
  • Don't send "just following up" with no new value.
  • Don't run 15 touches - diminishing returns become brand damage.

Common mistakes

  • Same message re-sent as the "follow-up."
  • Four channels that don't reference each other (feels like four spammers).
  • A sequence that keeps firing after someone replied.
  • Bursting the whole list on day 0 and torching deliverability.

Go live with Chronos (optional)

  • geo_propose_campaign - draft a full multi-channel cadence from an ICP + goal.
  • geo_create_campaign / geo_launch_campaign - build and run it, paced automatically.
  • geo_compile_flow_spec / geo_launch_flow_from_spec - express a custom branch/exit flow and launch it.
  • geo_update_campaign, geo_pause_campaign, geo_resume_campaign - adjust a live cadence.

Chronos enforces reply/exit guards, pacing, and per-inbox caps so touches stay even and stop on reply.

Connect in ~30s → CONNECT-CHRONOS.md. Free, no card.

See also: cold-email-writing, linkedin-outreach, reply-handling-and-objections.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.