Email Deliverability Preflight
Verify authentication, sender reputation, inbox placement, and list health before sending or scaling email.
The skill treats exports, DMARC reports, DNS data, and HTML as untrusted input, requires scope/profile confirmation, and asks before writing memory. Connectors are described as read-only by default, keys are not persisted, and external sending requires an explicit --live flag. Deducted 5 points because memory writes, DNS/ESP access, and optional external mutations still depend on host controls and shared protocols; least-privilege, rollback, and sensitive email-data handling are not fully demonstrated here.
Inputs, outputs, NEEDS_INPUT handling, abnormal cases, and termination conditions are unusually explicit. Deducted 12 points because static review cannot establish execution of this skill's key paths; scoring, handoff, and connector behavior depend on shared reference files not supplied here, and there are no skill-specific tests or reproducible failure examples.
Trigger phrases, four sending profiles, required inputs, non-fit ranges, and handoff boundaries are clear, with Chinese-language labels and a keyless manual path. Deducted 3 points because availability of Google Postmaster, Microsoft SNDS, seed-list testing, and DNS-over-HTTPS is not fully addressed; mainland-China reachability and substitutes have limited evidence, although the core flow can use user-provided exports.
The frontmatter declares name, slug, version, license, compatibility, triggers, and argument hints. Quick Start, Contract, Data Sources, Instructions, Save Results, and Next Best Skill provide good layering, while repository materials include Apache-2.0, README, CI, and tests. Deducted 4 points because maintenance ownership, changelog linkage, and update path are incomplete for this skill; SECURITY.md still lists 17.x as supported while the skill and repository identify 18.0.0, and several important referenced files were not provided.
The promised output is directly structured around per-item checks, S1 status, SEND-S scoring, profile naming, and handoff, and the checklist is usable for a manual pre-flight. Deducted 9 points because no representative output, execution result, or third-party validation is included; score computation depends on the absent send-benchmark, and users must supply and reconcile many heterogeneous data sources.
The skill specifies data sources, check items, and Pass/Partial/Fail/NEEDS_INPUT rules, and the repository includes general architecture and audit-validator tests. Deducted 6 points because the supplied tests do not demonstrate coverage of deliverability-qa key paths, and there is no real run evidence, independent reproduction, or corroborating third-party evidence.
- This is a static source review only; no scripts, DNS lookups, ESP checks, or seed-list tests were executed. Security and reproducibility claims in the documentation are not independently verified.
- Before use, confirm host isolation for memory writes, user-consent enforcement, and handling of sensitive email data; also resolve the 18.0.0 versus SECURITY.md 17.x support-version mismatch.
- The materials do not establish mainland-China reachability for Google/Microsoft reputation data, seed-list services, or DNS-over-HTTPS. When unavailable, use user-provided exports and retain NEEDS_INPUT status.
What does this skill do, and when should you use it?
Deliverability QA is the setup-stage email skill in the repository’s SEND framework. It runs a one-time pre-send review covering SPF, DKIM, DMARC, BIMI, sending-domain and IP reputation, inbox placement, spam-content and rendering risks, and point-in-time bounce and complaint rates. It reports pass, partial, or needs-input status for each sub-item, then produces the SEND-S score and S1 authentication flag. It does not calculate the full EQS or own recurring list-health monitoring, so it fits teams that have campaign evidence available and need a documented preflight before a send or scale-up.
Reads the sending domain, one typed profile, SPF/DKIM/DMARC/BIMI DNS records, the DMARC aggregate RUA report, a seed-list inbox-placement test, ESP deliverability and domain/IP reputation reports, and campaign HTML. It checks authentication presence, alignment, and pass status; evaluates inbox, spam, and Promotions placement; scans for spam-trigger language, image-to-text imbalance, broken or mismatched links, missing plain text, and rendering failures; compares hard-bounce and spam-complaint snapshots with benchmarks; and emits per-item findings, the SEND-S score, the S1 flag, missing evidence, and a handoff summary. With user approval, results can be saved under memory/email/deliverability-qa/.
- An email marketer is preparing a promotional or newsletter send and needs to verify SPF, DKIM, DMARC, and BIMI alignment.
- An ESP user is landing in spam or Promotions and wants to correlate seed-list placement with reputation and authentication evidence.
- A new sending program is warming an IP or preparing to scale and needs a one-time readiness snapshot.
- A marketing operator wants pre-send QA for email HTML, links, plain-text content, and rendering risks.
What are this skill's strengths and limitations?
- Covers authentication, reputation, placement, content/rendering, and point-in-time list-health evidence in one pre-send workflow.
- Clearly separates the S1 prerequisite and SEND-S score from the full EQS gate.
- Works with user-provided exports and manual DNS or seed checks without requiring paid tools or MCP.
- Treats missing evidence explicitly and flags failing authentication as a veto candidate.
- It is a one-time preflight, not a recurring monitor for bounce, complaint, or list-decay trends.
- The quality of the result depends on user-provided reports, DNS evidence, seed testing, and campaign HTML.
- It does not calculate EQS or enforce S1, S2, N1, or D1; those responsibilities belong to email-quality-auditor.
- The supplied material does not document an independent test suite or a complete list of tested ESP platforms.
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 a compatible generic Agent Skills host, run npx skills add aaron-he-zhu/aaron-marketing-skills. A direct clone is also supported: git clone https://github.com/aaron-he-zhu/aaron-marketing-skills. The repository is licensed under Apache-2.0.
How do you use this skill?
Prepare the sending domain, exactly one profile (promotional, retention, cold-outbound, or newsletter), a DNS export, DMARC RUA report, seed-list placement test, ESP deliverability/reputation report, and campaign HTML. Then use: "Run a deliverability pre-flight for [sending domain] before I send. Here is my DMARC RUA report, a DNS export, and my seed-list inbox-placement test: [paste/path]." Missing RUA or inbox-placement evidence is reported as NEEDS_INPUT rather than passed by default.
How does this skill compare with similar options?
Unlike list-hygiene-monitor, this skill reads a one-time pre-send list-health snapshot rather than recurring trends. Unlike email-quality-auditor, it builds and verifies the SEND-S/S1 signal; the auditor calculates full EQS and enforces the applicable vetoes.