draft-sequence
Skill1 starsChanged 4 months agoMITView on GitHub
Draft a complete multi-touch Outreach sequence — calls, LinkedIn, and emails — for a play or a target list. Use when you say "build a sequence", "draft the cadence for...", "what should the touches be", or after build-list produces a Tier A list.
In this repository14 other files it gives its agents.
- account-briefSkill
- battlecardSkill
- build-listSkill
- dojoSkill
- handle-replySkill
- meeting-prepSkill
- post-callSkill
- reactivateSkill
- setupSkill
- signal-watchSkill
- voice-interviewSkill
- what-worksSkill
- work-inboundSkill
- write-emailSkill
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 12
Read ${CLAUDE_PLUGIN_ROOT}/references/messaging.md and references/guardrails.md.
- Distinctive: line 19
Every email obeys the write-email skill's rules (≤80 words, one CTA, locked structure, banned words).
≈ “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.
Files
Draft Sequence
Build a full Outreach-ready cadence. Input: which hook/play it runs (see messaging reference), who it targets (persona + list), and any campaign tie-in (default: your current campaign from messaging.md, if one is live).
Before you draft
Read ${CLAUDE_PLUGIN_ROOT}/references/messaging.md and references/guardrails.md. Look for voice-profile.md in the working folder for call-script and email voice, and learnings.md (from the what-works skill) — build on its "Proven" patterns, never on anything "Killed."
Design rules (non-negotiable)
- Call-first architecture (guardrail #3 — email deliverability is compromised until your admin says otherwise). Calls and LinkedIn carry the sequence; email supports.
- 10-14 business days, 8-10 touches, then out. No zombie sequences.
- Max 2 emails per thread without a reply. Bump = reply-over in the SAME thread, 2 sentences (configure in Outreach as a follow-up step on the prior email, or "saw this" language breaks).
- Every email obeys the write-email skill's rules (≤80 words, one CTA, locked structure, banned words).
- Breakup email: never "bringing this back to the top of your inbox." Give a reason to keep the door open (the campaign's recordings/offer works: "Registration link either way — the recordings go to everyone who signs up").
Default skeleton (adapt, don't worship)
| Day | Touch | Notes |
| 1 | LinkedIn profile view + connect (no note, or 1-line trigger note) | warm the name |
| 1 | Call #1 + voicemail (15 sec: trigger + frame + name drop, no pitch) | |
| 2 | Email 1 — trigger/frame/trap/CTA | thread starts |
| 4 | Call #2 (no voicemail) | |
| 5 | Email 2 — same-thread bump, 2 sentences, campaign link (if live) | reply-over |
| 7 | LinkedIn message (if connected) — shortest version of the trap question | |
| 9 | Call #3 + voicemail #2 (different angle: proof point or campaign offer) | |
| 11 | Email 3 — new thread, different hook angle | |
| 14 | Breakup — one line + recordings/offer | out |
What to produce for each touch
- Emails: full copy per the write-email structure, with build notes for Outreach (thread behavior, merge fields as
{{first_name}}/{{company}}/{{trigger}}). - Call scripts: opener (trigger + reason for calling, 2 sentences) → permission ("30 seconds?") → frame → ONE discovery question → soft CTA (meeting or campaign offer). Plus a 15-second voicemail script. If a campaign is live, keep session lengths accurate to what's actually being offered.
- Objection branches (calls): "we use the status-quo tool / we have the legacy incumbent / no budget / not my job" — each answered with the matching differentiator from the messaging reference, 2 sentences max, always ending in a question.
- LinkedIn touches: ≤40 words, no links in the connect note.
Output format
# Sequence: [name] · Hook #[N] · persona: [type] · [date]
## Architecture
[the day/touch table, adjusted]
## Touch-by-touch
[Day N — type]
[full copy or script]
[Outreach build note]
## Objection branches
[objection → response]
## Before this goes live (checklist)
- [ ] Names/proof cleared with your marketing lead [⚠️ if used]
- [ ] Campaign details/session times verified (if a campaign is referenced)
- [ ] List deduped against HubSpot active pipeline
- [ ] Your admin or marketing lead reviewed the copyThat last checklist is not decoration — a sequence does not go live without all four boxes.
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
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.