Reactivation Specialist
Design a closed-loop win-back program for dormant email subscribers before they are suppressed.
The skill narrowly designs a reactivation program and does not directly send mail or own consent records. It treats exports as untrusted, excludes suppressed subjects, requires renewed opt-in, delegates recording to consent-registry, and requires user confirmation before saving results. Seven points are deducted because data minimization, sensitive-export handling, rollback/recovery, and the complete write data flow are not fully specified, and the compliance loop depends on another skill.
Triggers, inputs, outputs, completion criteria, stop conditions, and NEEDS_INPUT handling are fairly coherent; the ladder explicitly stops instead of looping. Two points are deducted because there are no skill-specific key-path tests, representative completed outputs, or reproducible failure messages. Static review caps this dimension at 10.
Audience, semantic triggers, exclusions, parameters, and expected outputs are explicit, with bilingual material and a keyless Tier-1 path. Four points are deducted because ESP field mappings, jurisdictional compliance variation, Chinese-market requirements, and mainland-China reachability are not concretely addressed; geo-relevance is marked low.
The skill has structured frontmatter, versioning, license metadata, quick starts, a contract, data sources, instructions, decision gates, save template, references, and termination rules; Apache-2.0 is supported at repository level. Five points are deducted for no skill-specific changelog, clear maintenance owner, or update path, plus a governance inconsistency between SECURITY.md supporting 17.x and the skill declaring 18.0.0.
The core task is decomposed into cohort definition, staged offers, re-permission, terminal suppression, and handoff facts, making the intended program directly usable in principle. One point is deducted because no representative output or outcome evidence is supplied, and consent recording, lifecycle reintegration, and final auditing remain dependent on other skills. Static review caps this dimension at 7.
The skill supplies a clear contract, source-data labeling rules, Measured/Estimated distinctions, and auditable handoff boundaries; the repository also shows CI and tests. One point is deducted because the supplied tests do not cover this skill's key paths and there is no third-party execution evidence or independent reproduction. Static review caps this dimension at 5.
- Do not treat an open or click as renewed consent; define the affirmative action, timestamp, and lawful basis, and record them through consent-registry.
- The skill designs suppression but does not perform the final audit; before sending, verify unsubscribe integrity, global send caps, complaint signals, and applicable law.
- Confirm that shared references, ESP export fields, and 18.0.0 version governance are consistent; repository-level tests do not establish that this skill itself is tested.
What does this skill do, and when should you use it?
Reactivation Specialist plans recovery and clean retirement for email subscribers who have stopped opening or clicking. It defines the lapsed cohort, builds an escalating offer ladder, adds a fresh re-permission action, and sets a terminal sunset-or-suppress rule. It fits email teams that need to recover valuable inactive subscribers while reducing risk from a decaying list. It does not design everyday lifecycle flows, own preference-center logic, store consent records, or calculate SEND quality scores.
Reads the stated no-open or no-click window, ESP engagement or flow exports, available incentives, complaint signals, suppression policy, and one SEND profile; defines how the lapsed cohort is selected and labels cohort findings Measured or Estimated; creates a staged ladder consisting of soft re-engagement, conditional incentive, and last-chance messaging with triggers, delays, intent, exit-on-engagement conditions, and stop rules; specifies a re-permission action such as click-to-confirm, a preference-center visit, or an outbound reply; routes re-permissioned subjects back to active nurture and sends non-responders to confirmed suppression after a defined window; produces a reusable handoff summary and, after confirmation, saves results under memory/email/reactivation-specialist.
- An email manager needs a win-back campaign for subscribers who have not opened messages for several months.
- A marketing team wants to recover valuable inactive contacts before applying a sunset policy.
- A compliance-conscious team needs re-engaged subscribers to explicitly confirm continued opt-in.
- Complaint rates are rising in the inactive tail, so the team needs capped frequency and a terminal suppression rule.
- A lifecycle team needs to hand re-permission and sunset facts to consent, nurture, and preference-management owners.
What are this skill's strengths and limitations?
- Covers the full planning loop: cohort definition, offer escalation, re-permission, and terminal suppression.
- Requires every subject to reach either a re-permissioned or suppressed state, preventing indefinite looping.
- Clearly separates re-permission design from consent-registry recordkeeping.
- Works from user-provided data by default; ESP APIs and MCP access are optional.
- Includes frequency caps, quiet-hour handling, and exclusion of already suppressed or hard-bounced subjects.
- Does not design welcome, cart-abandonment, browse-abandonment, or post-purchase lifecycle flows.
- Does not store consent or suppression records, calculate EQS, or enforce SEND auditor vetoes.
- Without engagement exports, cohort size and recency distribution remain estimates.
- The supplied source shows no platform-specific validation results or standalone test suite for this skill.
How do you install this skill?
The skill is distributed in the aaron-he-zhu/aaron-marketing-skills repository. Run git clone https://github.com/aaron-he-zhu/aaron-marketing-skills, then use email/nurture/reactivation-specialist/SKILL.md with an Agent Skills-compatible host. The README also documents Claude Code installation: /plugin marketplace add aaron-he-zhu/aaron-marketing-skills followed by /plugin install aaron-marketing@aaron. For generic hosts, it documents npx skills add aaron-he-zhu/aaron-marketing-skills. No single-skill installation command is documented in the supplied source.
How do you use this skill?
Provide the lapsed cohort window, ESP or platform, available offer, suppression policy, and any engagement export. For example: "Build a win-back campaign for subscribers who have not opened in 90 days on [ESP]. I can offer [incentive]. Define the re-permission step and suppress non-responders after the final window." If the engagement export is missing, the skill can proceed with estimated cohort findings; if the window is unknowable or consent records cannot be checked, it should request those inputs first.