B2B Cold Outbound Sequencer
Map safer B2B cold-email sequences, reply routing, mailbox warmup, and compliance guardrails.
The skill treats exports as untrusted input, requires lawful-basis checks, routes opt-outs to immediate suppression, keeps keyed connectors optional, and explicitly avoids sending mail or giving legal advice. However, memory promotion/save behavior, rollback, and sensitive business-data boundaries are underspecified, while attribution is mainly internal repository referencing, so 7 points are deducted.
Inputs, outputs, done conditions, NEEDS_INPUT behavior, missing-ESP handling, and the no-fallthrough reply rule are documented coherently. Static evidence provides no skill-specific tests, reproduced outputs, or diagnosable key-path failure traces, so 11 points are deducted under the static calibration cap.
Triggers, audience, parameters, expected outputs, and non-fit cases for B2C, consent records, and EQS computation are explicit. Jurisdiction handling remains high-level, Chinese support is mostly metadata and summary text, and environment-fit evidence is limited, so 4 points are deducted.
The document follows a consistent Quick Start, Skill Contract, Data Sources, Instructions, scope, and handoff structure, with version, Apache-2.0 licensing, metadata, limitations, and next-skill routing. Maintenance ownership, changelog, and update path are not clear in the skill, and version 18 conflicts with SECURITY.md's listed supported 17.x line, so 5 points are deducted.
The contract directly specifies a sequence map, reply-triage table, warmup/throttle plan, compliance guardrails, and SEND S read, making the core task plausibly usable. There is no representative finished output, execution evidence, or delivery-data validation; warmup figures may be Estimated and legal review remains necessary, so 9 points are deducted.
The skill identifies data sources and distinguishes Measured, User-provided, and Estimated evidence, but the referenced benchmark and shared protocol files are not fully included here, and repository tests do not cover this skill's key paths. Static verifiability is therefore limited, so 7 points are deducted.
- The skill plans cold-email outreach but does not verify live headers, consent records, or sending configuration; validate these with consent-registry, ESP/DNS data, and email-quality-auditor before sending.
- CAN-SPAM, GDPR/PECR, CASL, and similar jurisdictional statements are operational guidance only, not legal advice; do not assume approval without a recorded lawful basis.
- Warmup and daily-throttle figures may be Estimated and are not guarantees for a particular domain, mailbox, or industry.
- This was a static source review only; the skill, connectors, and save workflow were not executed, and reproducibility of key outputs is unproven.
What does this skill do, and when should you use it?
This is the Email/Deliver skill for designing B2B cold-outbound programs in Aaron Marketing Skills. It produces a multi-step sequence map, reply-triage branches, per-mailbox warmup and throttling schedules, and jurisdiction-aware compliance guidance. It reads the SEND S dimension, but does not compute EQS, own the consent record, adjudicate lawful basis, write email copy, or provide legal advice.
It reads the user's ICP or sequence goal, domain and mailbox setup, list source, target jurisdictions, and optionally an email-platform sending report. When available, it can use SPF, DKIM, DMARC, and DMARC aggregate-report information for the sender-integrity read. It defines channel, timing, step goal, and exit rule for each touch; routes positive, objection, referral, not-now, and opt-out replies; plans mailbox-level warmup and daily send caps; and lists CAN-SPAM elements such as accurate identity, non-deceptive subjects, a physical address, and a working opt-out. It emits a SEND S read with the Cold-outbound typed profile and a handoff summary, with a confirmed result optionally saved to the documented memory path.
- A sales team has an ICP and target list and needs the outbound sequence mapped before writing individual emails.
- An outbound operator needs deterministic routing for positive, objection, referral, not-now, and opt-out replies.
- A team has a fresh sending domain or several new mailboxes and needs a staged warmup and daily throttle plan.
- A marketer is targeting multiple jurisdictions and needs to flag places where a lawful basis may be required beyond an opt-out model.
- A program owner wants a structured handoff to consent tracking or the SEND quality auditor before sending.
What are this skill's strengths and limitations?
- Covers sequence structure, exit rules, reply routing, warmup, throttling, and compliance guardrails in one workflow.
- Explicitly requires hard exits for replies, opt-outs, bounces, and the natural end of the sequence.
- Works from user-provided data at Tier 1; email-platform APIs and MCP access are optional conveniences.
- Produces a focused SEND S read and clearly hands off consent and final scoring to the appropriate skills.
- Does not write subject lines, body copy, or CTAs; those belong to email-creative-builder.
- Does not store or adjudicate consent and lawful-basis records; missing evidence becomes NEEDS_INPUT.
- Compliance content is operational guidance, not legal advice, and requires counsel confirmation.
- Without measured warmup data, ramp volumes are category-standard estimates rather than account-specific recommendations.
- Does not calculate EQS or enforce SEND S1, S2, N1, and D1 vetoes.
How do you install this skill?
For Claude Code, run /plugin marketplace add aaron-he-zhu/aaron-marketing-skills, then /plugin install aaron-marketing@aaron. On other Agent Skills-compatible hosts, run npx skills add aaron-he-zhu/aaron-marketing-skills, or clone the repository with git clone https://github.com/aaron-he-zhu/aaron-marketing-skills. The README does not document a separate package command for this individual skill, but it supports the repository's generic single-skill retrieval approach.
How do you use this skill?
After installation, provide the sequence goal or ICP, sending domain and mailbox setup, target jurisdictions, and list source. Example: Build a 5-step cold-outbound sequence for [ICP] from [sending domain]. Here is my target list source and its jurisdiction mix: [paste/path]. You can also request: Design reply-triage branching for my outbound: route positive / objection / referral / not-now / opt-out to the right next action.
How does this skill compare with similar options?
It is distinct from email-sequence-designer, which handles consented B2C lifecycle flows; consent-registry, which owns consent and suppression records; and email-quality-auditor, which calculates EQS and applies the SEND gate. Individual email copy should be produced by email-creative-builder.